Please stand by. We're about to begin. Good day, welcome to Chipotle's third quarter 2016 earnings conference call. Today's conference is being recorded. At this time, I'd like to turn the conference over to Mark Alexee, Investor Relations Manager for Chipotle Mexican Grill. Please go ahead, sir.
Thanks, Don. Hello, everyone, welcome to our call today. By now you should have access to our earnings announcement released this afternoon for the third quarter of 2016. It may also be found on our website at chipotle.com in the investor relations section. Before we begin our presentation, I will remind everyone that parts of our discussion today will include forward-looking statements as defined in the securities laws.
These forward-looking statements will include statements of our business outlook, forecasts of EPS, comparable restaurant sales, restaurant-level operating margins, and G&A and other cost savings in the fourth quarter 2016 and for the full year 2017, and descriptions of the impacts of new technologies on our business, statements about planned marketing programs, projections of the number of restaurants we intend to open and new restaurant development costs, projections of effective tax rates for 2016 and 2017, and statements about stock repurchases, as well as other statements of our expectations and plans. These statements are based on information available to us today, we are not assuming any obligation to update them. Forward-looking statements are subject to risks and uncertainties that could cause our actual results to differ materially from the forward-looking statements.
We refer you to the risk factors in our annual report on Form 10-K as updated in our subsequent Form 10-Qs for discussion of these risks. I'd also like to remind everyone that we've adopted a self-imposed quiet period restricting communications with investors during that period. The quiet period begins on the first day of the last month of each fiscal quarter and continues until the next earnings conference call. For the fourth quarter of 2016, it will begin December 1st and continue through our fourth quarter for an earnings release planned for late January 2017. Our discussion today will also include non-GAAP financial measures, a reconciliation of which can be found on the presentation page of the investor relations section of our website. We will start today's call with some longer than normal prepared remarks, then we'll be sure to allow 20 to 30 minutes for questions.
On the call with us today are Steve Ells, our Chairman and Co-Chief Executive Officer, Monty Moran, Co-Chief Executive Officer, Mark Crumpacker, Chief Marketing and Development Officer, and Jack Hartung, Chief Financial Officer. With that, I'll now turn the call over to Steve.
Thanks, Mark, good afternoon, everybody. We are beginning to emerge from the most difficult year in our history and are seeing many reasons to be optimistic that we are headed in the right direction to restore our business to a place our many shareholders will be proud of. Please know that while we are on the road to recovery, we're not satisfied and will continue to work extremely hard to make the necessary adjustments necessary to restore our business and deliver results as quickly as possible. From early this year, we have recovered more than 16 points of sales comp, we've recovered 20 points of traffic comp, we're making steady progress toward further recovery. During the third quarter, we generated revenue of $1.04 billion, down 15% from last year on a comparable restaurant sales decline of 21.9%.
During the third quarter, we opened 55 new restaurants, bringing our total number of restaurants to 2,178. We generated diluted earnings per share of $0.27, which includes a negative $0.29 impact related to asset impairment. Since our recovery began, we have seen a gradually improving outlook for sales and earnings and fully expect that those trends will continue. 2016 will be marked as a significant reinvestment year as we continue to work to make Chipotle a stronger company in the wake of last year's events. Looking forward to 2017 and based on current trends, we believe that we will produce diluted earnings of $10 per share on same-store sales growth in the high single digits for 2017. We believe that the full year 2017, we can generate restaurant level operating margins of 20% and save more than $100 million in capital and operating expenses combined.
To achieve these results, we must be steadfast in executing our recovery plan, which remains the primary focus of everybody at Chipotle. Specifically, we're focused on delivering a safe and extraordinary guest experience in every restaurant, restoring trust and building sales, restoring our economic model, and enhancing the guest experience through innovation. Monty, Mark, and Jack will provide additional detail on some of these priority areas, right now I'd like to speak to the last point, which is enhancing the guest experience. Two important enhancements to the customer experience are menu innovation and digital ordering. We are actively exploring ways to enhance the guest experience by offering new additions to our menu. We believe that this is a good way to entice infrequent or lapsed customers to return, as well as a way to increase sales.
Anytime we consider new menu items, it's important to us that they remain consistent with our food culture, both in terms of the ingredients we use and how the new items are prepared. A great example of this is chorizo, which we launched nationwide earlier this month. Chorizo now accounts for roughly 7% of entree sales around the country. Feedback from customers has been extremely positive, and we plan to ramp up our advertising in the fourth quarter to support chorizo by encouraging our customers to come in and try it. We're also currently testing two different desserts, and we hope to select one to offer in the near future. We believe that there's an opportunity to excite current customers and to attract new customers through thoughtful menu development, and we're exploring these options.
We're also moving aggressively to make digital ordering more appealing to our customers and more efficient for our restaurants. Nearly all of our restaurants have a second make line in the back of the restaurant where we fulfill digital, fax, and catering orders. By optimizing the use and design of these second make lines, we envision a time when digital ordering could account for a much larger percentage of sales than the 6% it represents today. Maximizing our digital sales requires changes both to the second make line itself and to the ways by which our customers place digital orders. Our original second make line resembles the service line our customers see in the front of the house, though scaled down, and it takes up less space.
The new line is a much more ergonomic and efficient design configured such that less labor will be required to produce higher sales volumes, and does so more quickly. The new line incorporates an advanced queuing system that takes advantage of a heads-up visual display, which allows our crews to assemble orders without distraction. The benefits of these new lines will be enormous, including faster, more efficient assembly of orders, reduced errors, and more consistent portion sizes for our customers, all of which we believe will help us drive increased sales volumes. After some additional testing is complete, we will begin to selectively add these new digitally enabled lines to some of our highest volume markets where it can best enhance our out-of-store ordering experience.
It's important to note, however, that all of our original second make lines are still able to take advantage of enhanced digital ordering options, even though they don't have the advanced indicating system of the digitally enabled lines. Our crews in these restaurants will work from paper tickets until their restaurants are upgraded with the new system. To stimulate additional use of these second lines, we're introducing a new mobile-optimized ordering site that provides an alternative to our iOS and Android ordering apps. Mark will talk more about this new ordering option later in the call, but the new site makes ordering and paying for digital orders easier than ever.
In addition to our new ordering site, we are testing a tablet-based ordering option, which allows customers to avoid the line and order in-restaurant on available tablets and get their order made from the second make line in the back of the house. This tablet-based ordering system will initially be tested in a series of restaurants that are equipped with the new make line. When the full system is realized, the second make line has the potential to deliver as much as 50% of the sales from some locations, including digital orders from all sources, including catering and delivery services. Finally, we fully impaired the assets related to ShopHouse, which total approximately $14.5 million on a pre-tax basis. We opened the first ShopHouse in 2011 and currently operate 15 ShopHouse restaurants in three separate markets.
ShopHouse is managed by a small independent team, which was tasked with creating a compelling brand which could develop into a compelling growth strategy and enhance long-term shareholder value. We're proud of the ShopHouse team, the delicious food they created, and the excellent customer service they've delivered. After operating in three distinct markets and opening in a variety of trade areas, ShopHouse simply has not demonstrated the ability to support an attractive unit economic model. As a result, we have decided not to invest further in developing or growing the ShopHouse brand and will pursue strategic alternatives.
Although the exotic ShopHouse cuisine was not able to attract sufficient customer loyalty and visit frequency to make it a viable growth strategy for us, we continue to believe that our approach to food, people, and unit economics with the right cuisine, with the right concept, can lead to a compelling growth strategy. We are optimistic that our other growth seeds, serving pizza and soon burgers, both of which have broad customer appeal, can become further growth strategies for us. While this was not an easy decision, we trust in our overall growth seed approach, where we can innovate and develop new concepts with a relatively small capital investment and little or no distraction to the Chipotle business. We'll only invest additional resources, both time and money, into growth seeds that show the potential to become meaningful growth vehicles and a compelling way to enhance shareholder value.
We continue to believe that we have an extraordinary potential and are fully confident in our ability to achieve it. Monty, Mark, and Jack will talk in greater detail about our plans to realize our potential and build greater shareholder value. I'll now turn the call over to Monty.
Thank you, Steve. Achieving a full recovery means having all of our restaurants at their very best in terms of serving delicious, safe food and providing an excellent restaurant experience while continuing to grow sales. As we move forward with our recovery plans, the guest experience remains our greatest area of focus. Providing an excellent guest experience starts with serving food that's always safe and delicious. While many of our enhanced food safety efforts occur outside the restaurants in the supply and distribution network and with enhanced technology, we also rely on strong restaurant teams and field leaders to carry out additional elements of our food safety program in the restaurant. Ensuring that we execute these procedures in our restaurants has been our top priority. To accomplish this, we have significantly expanded training, food safety inspections, and third-party and internal audits in our restaurants.
These audits help us identify opportunities for improvement and ensure that all of the programs we have put in place are being executed consistently and properly. To further these efforts and to continuously improve food safety in our restaurants and supply chain, we have also established an independent food safety advisory council made up of some of the country's foremost experts in food safety and food microbiology. This advisory council, alongside Dr. James Marsden, our Executive Director of Food Safety, is charged with continuously reviewing our food safety programs and looking for opportunities to strengthen them even more. This quarter, we shared some key points about our enhanced food safety program with our customers through a PR campaign aimed at highlighting eight key areas of advancement that we've made.
The campaign included full-page ads in newspapers as well as online and social media advertising, in addition to print and video materials of Steve describing why our food safety changes are so effective. The campaign was very well received and was effective in communicating the comprehensive steps we've taken. Details surrounding each of these programs continue to be available for you to see on our website. At the same time, we are keenly aware that safe food alone will not bring people into our restaurants. Instead, they come for delicious food and an excellent guest experience. This is why our restaurant teams are focused on delivering great-tasting food and excellent customer service in clean, friendly restaurants. This starts with our commitment to serving the best-tasting food we can, food that is made with ingredients that are raised with respect for the environment, animals, and the people who produce them.
We continue to look for ways to improve the quality of the food we serve, including our quest for better tortillas made with no artificial additives or preservatives, a promise we made last year and are close to fulfilling. As we work to rebuild our company and strengthen our brand, it is critical that our entire organization is focused on those things that will most powerfully delight our guests and speed our recovery. For this reason, we used this year's All Managers Conference as a time to articulate to our field leaders and managers three areas of focus that are in their control. First, providing an excellent guest experience. Second, marketing their individual restaurants. Third, restoring and protecting our unit economic model. Our officer team asked our managers to first train the fundamentals of excellent operations so that they could then proceed on the path to Restaurateur.
Our relentless focus on the most important things and implementing them exceptionally well is what has always made Chipotle great. We know that to ensure our sales recovery, we have to do better than ever to win our customers back. Also, great throughput is a key part of the guest experience at Chipotle and one of our key competitive advantages. To do it well requires constant training and focus. During the last year, where we made implementation of our new food safety program the highest priority, some of our restaurant teams have lost some of the throughput skills that they had developed before the crisis. With high turnover at the crew level, it is even more important that we continue to emphasize and train the skills needed to deliver fast throughput, including the four pillars of throughput.
Of course, a primary reason our throughput slowed was that our sales slowed after the crisis. Throughput still can be and will be faster as our restaurant teams commit to drive this critical aspect of an excellent guest experience. Additionally, we are emphasizing the proper use of the second make line with our field teams to enable the teams on the front line to take great care of the guests in front of them. This will be even more important as we layer in some of the technology that Steve spoke about, which will revolutionize our use of this second way of serving our guests. Our initiatives surrounding catering and technology adoption for mobile and online ordering round out the makings of an excellent customer experience and will allow for further sales recovery and future growth.
Our Restaurateur program remains central to our ability to continue to produce the leaders we need to accomplish our mission and guide our restaurants through the sales recovery. This program has allowed us the ability to relocate some of our best operational leaders in order to establish better uniformity of operations around the country. Specifically, in our fast-growing and high-volume Northeast region, which includes New York and New England, we were not satisfied with our guest experience in terms of throughput and dining room cleanliness. By relocating some of our strongest executive leaders directly into these markets, we are now confident that they will quickly return these restaurants to the operational excellence that we and our customers have grown to expect. These leadership changes include the relocation of an executive regional director and an executive team director, as well as the addition of five team directors.
We have confidence that our newly strengthened Northeast team will provide much-needed support to our general managers and quickly and dramatically improve the guest experience. Over the last four years, we've added many field leaders to our ranks to be sure that we have the right ratios of restaurants per leader to carry out our mission. I've met with all of our field executives since then. I'm confident that they are all clear on our priorities and that they're the right people to help us achieve our mission. Prior to last year's issues, Chipotle had the strongest economic model in the industry. Of course, this model has been weakened due to lower sales volumes that we've seen this year.
While it's critical to fully restore sales volumes and keep improving them from there, we also know that we need to improve our economic model now so that we can provide healthy returns even at lower volumes. Key to restoring our business model is making sure that we are being as efficient as we can with our financial resources, but doing so without compromising the strength of our support teams and field leadership. We are using this crisis as an opportunity to take a critical and disciplined look at how we manage our financial resources, and we've established a goal of saving $100 million across operational, G&A, and CapEx expenditures. These cost savings will not detract from the customer experience.
We're finding these savings in a number of areas, including disciplined staffing and support departments and field leadership, better business terms with suppliers, operational efficiencies, and new restaurant designs that remain true to our brand but have lower investment costs. This more disciplined approach to managing expenses will allow us to strengthen our economic model to the benefit of shareholders while not compromising our ability to deliver an excellent customer experience. We continue to believe that expanding Chipotle in Europe holds a lot of promise for our future, we know that to best ensure our success there, we need to provide the area with more strategic attention. For this reason, we recently hired as managing director in Europe, Jim Slater, a proven industry leader who ran successful operations for Costa Coffee and led Costa to become the number one coffee brand in the U.K.
Jim will be working to develop our European markets into an investable growth strategy. We're excited to have Jim on board leading our European team, together they will work to determine the best strategy for our continued development in Europe. Through all of these initiatives, we are confident that the plan we have in place to drive our recovery is the right plan, we know that we have proven leaders throughout this company who can inspire a culture of excellence and help lead the way to recovery. I'm confident that through our progress towards providing a better guest experience all of the time, increasing sales through marketing, and strengthening our unit economic model, we will dramatically improve returns for our shareholders in the months and years ahead. I'll now turn the call over to Mark.
Thanks, Monty. I'm very glad to be back, I'm eager to share with all of you some of the important work that we're doing to bring customers back and restore our brand image. First, I want to apologize to everyone for recent events in my personal life. I'm sorry that I caused a distraction for the company, I want you all to know that I've put this behind me, learned from it, returned to my role in early September, excited and with a renewed focus and determination to help drive Chipotle's recovery. I want to thank you all for your support and forgiveness, which has meant a great deal to me. Today, I'm going to recap the marketing activities from Q3, then I'm going to talk about several important breakthroughs with regard to customer data, technology, and restaurant design before reviewing development expectations.
During the quarter, we successfully concluded the Chiptopia Summer Rewards program, designed to restore the frequency of our most loyal customers. In all, we had just under 6 million participants, with more than 2.5 million of those earning rewards in the program. We issued 340,000 super rewards for customers who maintained higher frequency levels, including more than 75,000 Catering for 20 rewards. We're excited by the potential positive brand impact of 75,000 catering parties over the coming 6 months. Throughout the course of the promotion, we saw increased transaction and frequency levels, most important, we have nearly returned to pre-crisis levels among our most loyal customers. After the completion of the program, we anticipated traffic to fall slightly, we have instead seen our improved sales levels generally continue to hold, which is very encouraging. During the quarter, we ran three marketing campaigns, each with different but complementary objectives.
The first was the food safety advancements campaign Monty referred to, which was designed to communicate what we have done to ensure the safety of our food. The campaign was broad-reaching with more than 90 million impressions. Lapsed customers engaged more than any others with the video component of the campaign. The results of the campaign are positive, we will continue to communicate the benefits of our food safety program so that every customer can rest assured that our food is as safe as possible. The second campaign surrounded our Love Story animated short film, which was distributed through digital media, social channels, and in movie theaters across the country. The film has been viewed more than 60 million times, has proven to be highly effective at building trust in the Chipotle brand.
The campaign also included the Love Story game, which challenges players to match fresh ingredients while avoiding fake ones. The game launched earlier this month, quickly surpassed the initial results of our popular Friend or Faux? online game from the summer of 2015. Millions of people have played the Love Story game, more than 3 million BOGO rewards have been earned by playing the game. More than 35% of those have already been redeemed. The third campaign running during Q3 was the Ingredients Reign campaign, which is designed to increase visits to Chipotle and build brand trust by focusing on the high quality of our delicious ingredients. Ingredients Reign is running across nearly all media channels, including outdoor, radio, print, digital, social, and theater.
Both Love Story and Ingredients Reign campaigns exceeded our expectations, for this reason, both campaigns have been extended to run through much of the rest of the year. Finally, with regard to Q3, I'd like to provide an update on our latest consumer research. Admiration scores and serves healthy and nutritious food attributes have nearly returned to pre-crisis levels. Agreement that we have appropriately addressed the food safety issues increased to 90%. Sentiment on social media is increasingly positive, press mentions were 80% positive last month. Both are at or near pre-crisis levels. Consideration still lags pre-crisis levels by more than 10%, though it is up slightly for the quarter. Moving on to Q4 and beyond, we'll be leveraging significant improvements in our customer data capabilities along with a shift in marketing strategy to aggressively target new and lapsed customers.
By combining transactional information with other customer data, we are now able to identify more than half of our customers and reach them with specific offers and tailored messages. This important breakthrough not only allows us to target messages and offers to current and lapsed customers, but it also allows us to accurately measure the effectiveness of those efforts by tracking return visits. We have already delivered 2 targeted campaigns using this data. We are continuing to refine our programs in real time. This new customer data has implications well beyond more targeted and measurable marketing. The data allows us to very accurately measure customer frequency and details about their orders and behavior. For example, using this data, we now know that during the last 6 months, we saw nearly 30 million new customers at Chipotle.
These new customers were customers who we have not seen since before the food safety events of last year. In fact, transactions from these customers accounted for nearly half of all transactions over the last 6 months. Additionally, we know how many of these customers returned to Chipotle for additional visits and how many of them became regular customers. When combined with our marketing and advertising efforts, this detailed view into frequency and behavior allows us to evaluate the return from various marketing and promotional activities. Beyond customer data, we have also now begun to benefit from the strengthening of our technology infrastructure. In just a matter of days, we are launching a new online ordering experience that makes ordering and paying for Chipotle from any mobile device easier and faster. This mobile-optimized website works on any recent web-enabled device. It requires no app download.
The new ordering experience offers a beautiful and intuitive way to order from Chipotle. This new ordering experience is an addition to our existing apps for iOS and Android that currently have a combined user base of more than 2.6 million people. Additionally, our test of Smarter Pickup Times has concluded. We are in the process of rolling this technology out nationwide. Smarter Pickup Times allow us to adjust the wait time between order placement and pickup times based on the demand and capacity of a given restaurant at a given time. This means that the customer is always presented with the shortest possible time between order placement and pickup. Smarter Pickup Times works with our iOS and Android apps, as well as our new online ordering.
When combined with the advanced second make line Steve described earlier, has significant potential to increase our online ordering volumes and improve the customer experience. We expect Smarter Pickup Times will be rolled out nationwide by January. Recent advancements in our technology infrastructure are also going to have a significant impact on the customer experience of ordering catering at Chipotle. Next month, we will enable online ordering and payment for catering, which previously was only available via phone and required in-store payment. The ability to order and pay for catering online will dramatically improve the customer experience while simultaneously decreasing our call center costs. The next step with catering is offer delivery. Tests are currently underway. Both catering and mobile ordering will receive significant marketing support during Q4, a time of year where we typically see an increase in catering volume ahead of the holiday season.
We have also been supporting the national rollout of chorizo, and we're excited by the results. As Steve mentioned, chorizo currently accounts for about 7% of sales, and customer response has been very positive. 84% of customers who try chorizo like it, and 70% of those who try it say that chorizo increases their desire to visit Chipotle, and 77% say that the new offering makes them feel more favorably towards Chipotle. Because of these strong results, we plan to increase the reach of our chorizo advertising campaign by more than 400%. Next, I'd like to talk about television advertising and the test that's currently underway. We recognize the need for wide-reaching efforts to invite new customers to Chipotle. The fact that nearly 30 million new or lapsed customers visited Chipotle over the last six months shows that customers are responding to our advertising and promotional efforts.
In order to accelerate that momentum, we are considering national television advertising. Because our Ingredients Reign campaign tested positively in terms of customer perception and brand trust, we are running a 30-second ad from that campaign in multiple test markets. If the test reveals that television advertising delivers the results we want, we may begin to integrate national television buys into our advertising campaigns. You may also be aware that we are currently undertaking a review of advertising agencies. While we are pleased with our current agency partners, we have developed a robust new marketing strategy for 2017, and in order to ensure that we have the absolute best creative work on that strategy, we're reviewing potential new partners. The review will be complete by the end of the year and in time for the launch of a new campaign in the spring of 2017.
Going forward, we believe that there will be opportunities to incorporate new menu items into our overall marketing. We have been encouraged by the response we have seen with chorizo and see an opportunity to build excitement among existing customers and attract new customers through the thoughtful addition of menu items that are consistent with our brand. Any new menu item will be accompanied by extensive advertising campaigns designed not only to tout the new menu additions but to excite customers about coming into Chipotle. Finally, I'd like to report on the successful launch of a new restaurant design that has been in development for more than a year. The new design is an evolution of the current design, with improvements in lighting, acoustics, seating, customer flow, and the presentation of our kitchen.
Additionally, the new design is more cost-effective, with an average cost of $760,000, which is about $40,000 less expensive than our current restaurant design. We also anticipate a reduction in maintenance and repair as a result of this new design. This new restaurant design is already open in certain locations and under construction in several others. We expect that more than 50 new locations in 2017 will utilize this design, and even more than that will benefit from some aspects of the new design. Our new restaurants have been performing at approximately 73% of our comp restaurant volumes, slightly lower than pre-crisis levels. In areas where we have increased marketing of our new restaurant openings, that performance has improved, and we are closing the gap to those pre-crisis levels. In total, we are planning to open between 195 and 210 new restaurants in 2017.
As our returns on new restaurants have compressed this year, along with our margins, we've lowered the number of restaurants we plan to open. We will continue to focus on opening up the best potential sites that combine a lower risk profile with the highest possible returns. Even with our lower sales levels in 2016, we are not seeing a marked increase in cannibalization from opening new restaurants, which remains at less than 1% nationally. Most important, we view this reduction in new restaurant openings as temporary. Our long-term development pipeline remains robust, and with our new lower-cost restaurant design, we are setting our new restaurants up to achieve even higher levels of return on investment. We strongly believe that the ultimate potential for Chipotle restaurants in the U.S. could total 5,000. With that, I will now turn the call over to Jack.
Thanks, Mark. When the year began, we described 2016 as a year where we will make significant investments to ensure our food is safe and encourage our customers to return to Chipotle. We knew these significant investments would put additional pressure on our margins in the form of elevated food costs, higher advertising and promotion costs, and higher labor costs to serve our returning customers. We intentionally did not emphasize restaurant efficiencies and even encouraged extra staffing in our restaurants during promotion peaks, because we did not want to risk trading a few points of margin on the P&L in return for disappointing a customer who visits an understaffed restaurant. The sales decline, combined with these investments and inefficiencies, have resulted in significant declines to our margins.
As we near the end of 2016 and we begin to look ahead to 2017, we'll continue to do all we can to restore customer trust and recover customer visits. We also are committed to taking important steps to restore our economic model. With the strategies you've heard Steve, Monty, and Mark talk about, we believe we can deliver a comp in the high single-digit range for the full year in 2017. Of course, we hope these strategies will deliver an even greater sales increase, we believe this comp projection represents a stretch, yet responsible goal.
With a focus on delivering greater efficiencies in our restaurants, identifying cost savings in everything we buy in our restaurants and in our offices, supporting our growing restaurant base with flat G&A, and beginning to normalize our promotion and advertising, we expect to deliver a restaurant level margin of 20% and an EPS of $10 in 2017. While these projected results do not fully restore our economic model, it's an important step in the right direction. We continue to believe we can, over time, fully recover our restaurant margins as sales recover. From a unit economic standpoint, these projected results would lead to an average restaurant volume of around $2 million and restaurant ROI of around 50%. The return potential for new stores is even greater, as we expect the average new store investment to decline to around $760,000 in 2017.
I'll talk more about what to expect in 2017 and in Q4. Now let me recap our results for the third quarter. During the third quarter, we reported sales of $1.04 billion, a 15% decrease from the third quarter of 2015. Our reported sales were reduced by $11.5 million because of deferred revenue related to our summer Chiptopia promotion. As a reminder, Chiptopia ended on September 30th. We recorded deferred revenues associated with rewards earned but not yet redeemed. These rewards can be redeemed or will expire over the next six months. The deferred revenue will be recognized over this period. As a result, we reported a negative same-store sales comp of -21.9%, which included a -0.8% impact from the revenue deferral. The Q3 comp is comprised of July, down 23.8%. That includes 100 basis points from the revenue deferral.
August was down 21.7%, which includes 70 basis points from the revenue deferral. September was down 20.1%, which includes 70 basis points from the deferral. While October comps for the first two weeks were affected by Hurricane Matthew and were down about 21%, the past 10 days have recovered and normalized and were down around 19%. While we're not satisfied with still being down around 19%, we continue to see small but steady progress month after month. As a side note, July had improved to declines in the 21% range in the weeks before our Q2 earnings report. Then was impacted by a negative month-end trading day effect, plus comparing to our Friend or Faux? from 2015, which ran the last week of July and into early August and was also impacted by the Chiptopia revenue deferral.
Our traffic trends have also steadily improved. Were down 15.2% overall in the quarter, compared to down 19.6% in June. Transactions improved to 13.4% in September. Most recently are running down 14% in the past week or so. Chiptopia helped drive the sales and transaction improvement in the quarter as our most loyal customers returned to their historic frequency patterns, as Mark mentioned. On a regional basis, we continue to see wide variances in the progress of the recovery, with comp sales in the Midwest and middle regions of the country down in the 17% range in September, outperforming the coastal areas, especially in the Pacific Coast and the Northeast, which are down closer to the 24%-25% range in September. From a transaction standpoint, there are several markets, including Kansas City, Chicago, Denver, and Minneapolis, where traffic is now down only in the high single digits.
We know that restoring customer habits takes time. To help ensure that renewed habits established during Chiptopia for our most loyal customers continue to hold, the Love Story campaign that Mark mentioned was launched this month. Although hard to estimate customers' behavior in October from only a few weeks of data, far our high-frequency customers are continuing to visit this month with similar frequency to September. Restaurant level margin was 14.1%, down from 15.5% in the second quarter of this year. The deferred revenue related to Chiptopia negatively impacted our restaurant level margin by nearly 100 basis points. Our underlying margin was 15.1%. The remaining decline from Q2's margin of 15.5% is due to higher marketing and promo costs related to Chiptopia and higher food costs. Food costs were 35.1% of sales. Excluding the effect of the deferred revenue, underlying food costs were 34.7%.
This represents a sequential increase from Q2 of 50 basis points, primarily due to much higher costs for avocados. Avocado supply declined during the summer. We experienced higher pricing, although we had hoped that this would be a temporary spike. In recent weeks, supply has become even tighter, and pricing has become much more volatile than expected. In fact, we've seen that some competitors recently have posted signs on their doors saying they're out of avocados altogether. We have remained in supply of avocados, but case pricing has risen from an average of about $30 per case during the first half of the year, to case pricing approaching $80 this month. Because of these higher avocado prices, food costs will remain above 35% in Q4 before seeing relief in early 2017. Our labor costs for the quarter were 27.6% of sales, up from 22.2% of sales last year.
Without the revenue deferral, labor costs were 27.3% of sales. Sales to leverage accounted for approximately 300 basis points, and the remaining increase relates to inflation and additional labor related to Chiptopia and other promotions. Underlying wage inflation was in the 5%-6% range in the quarter. Labor was about $9 million higher in Q3 versus Q2 this year to support higher sales levels as our transaction trends continue to improve. We also had increased labor as we return to prepping produce inside of the restaurants. We also had increased labor to support the higher promo level activity of Chiptopia. In prior years, total labor dollars had been relatively flat from the second quarter to the third quarter. Other operating expenses as a percent of sales were 16% during the quarter, an increase from 11.1% last year.
About 290 basis points of the year-over-year increase is due to sales through leverage. Marketing promo expenses combined were 4.8% of sales during the quarter, which is double the rate of marketing and promo from the prior year, and it was 50 basis points higher than Q2 due to Chiptopia. In Q4, we anticipate our marketing and promo will remain at about the same level as we test TV in a few markets. As we invest in ad supporting chorizo, and as the Chiptopia earned awards are redeemed and expensed in our promo line. G&A costs for the quarter were $78.4 million, or 7.6% as a percent of sales. Underlying G&A, excluding stock comp, was $61.9 million in the quarter. During the quarter, we held our biannual All Managers Conference, which totaled about $10 million, and $8 million of that was expensed in the third quarter.
As a reminder, this conference will not be held in 2017, although we traditionally hold a smaller field leadership conference, which we expect would cost between $1 million and $2 million. Excluding this year's AMC, our G&A was flat year to date on a total dollar basis compared to 2015, despite supporting an additional 247 new restaurants opened over the past year. We froze our support headcount early in the summer and have been actively prioritizing where we invest our support to ensure that we're investing in providing an excellent guest experience and recovering sales. As we look ahead to 2017, we are planning to continue to keep our underlying G&A expenses before stock comp flat compared to 2016, and lower on a percent of sales basis, despite growing our restaurant base by another 200 restaurants.
For the full year 2016, we estimate our effective tax rate will be 38.2%, in line with 2015. The full year 2016 estimated effective tax rate benefited from additional federal tax credits earned and a tax benefit from previously earned California state tax credits that we now estimate will be realized, they were offset by higher estimated state tax rates. We currently anticipate that the full year 2017 effective tax rate will be 39.5%, its higher expected rate for 2017 does not include the benefit of the previously mentioned tax credits related to prior years. Third quarter net income was $7.8 million, diluted earnings per share was $0.27. This included the negative impact of $0.23 related to the revenue deferral and a negative impact of $0.29 for the impairment of ShopHouse.
Before these adjustments, Q3 EPS was $0.79, or $0.08 lower than Q2 of this year, due to the AMC cost, the cost of Chiptopia, and higher avocado costs. As of September 30th, we maintained $609.8 million of cash and investments. During the quarter, we generated cash from operations of $101 million and generated free cash flow of $36 million. We also repurchased $70.3 million of our stock during the quarter at an average price of $409 per share. Over the life of our share repurchase program, we've repurchased $1.9 billion of our stock at an average price of $289 per share. We'll continue to manage our balance sheet to invest in our sales recovery, grow our business, and opportunistically return cash to shareholders via share repurchases.
Recently, our board of directors approved an additional $100 million of repurchase authorization, as of yesterday's close, including the additional approval, we now have $154.6 million remaining on our share repurchase authorization. Looking ahead to Q4, we will begin to compare against last year's comp of negative 14.6%. Of course, our comp will improve against that soft comparison and will likely be in the negative low single digits. Margins are expected to remain in the mid-teens as the benefit of a continued steady sales recovery and the reversal of the deferred revenue from Q3 will be offset by Q4 having seasonally lower average daily sales. G&A will be about $10 million lower than Q3 because of the AMC expense not recurring and slightly lower stock comp. We believe EPS will improve to around $1 per share.
Avocado pricing is the big wild card, as every $10 per case price move in up or down will impact EPS by $0.17. As I mentioned earlier, 2016 has been a year of significant investment in food safety, in restoring customer trust, and recovering customer visits. It has been a difficult year in terms of margins and earnings, driven by lower sales, including related to leverage, and the impact from these significant investments. We will continue to work harder to recover our sales, shifting our strategy from giveaways, discounts, and rewards to new menu items, technology-driven convenience, better operations including faster throughput, and more aggressive brand marketing, including TV. We know that fully restoring our previously industry-leading economic model depends on fully recovering our sales.
We feel confident that we have the teams in place to execute our strategies in 2017 and achieve a high single-digit comp sales increase, a 20% restaurant level margin, and a $10 EPS. Our comp target will not be an easy hurdle to meet, and of course, we are not in complete control of it. We are in control of how we manage our restaurants. We are in control of how we manage our expenses, both in the restaurant and with our support G&A. To earn a full year 2017 target of 20% restaurant level operating margin, we need to earn about 50 basis points of additional margin versus the current underlying run rate of about 15%. We will recover this 500 basis points in several ways.
First, by effective scheduling of our crews to ensure we have the right staff at the right times to deliver an excellent guest experience, but also to ensure we're not ever overstaffed, especially during low sales periods. We can manage our food costs more effectively by reducing waste, by ordering, prepping, and cooking the correct quantities throughout the day, by cooking to ideal temperatures, which not only makes our food taste better, it also results in avoiding unnecessary yield loss. We can also more effectively manage smaller but important restaurant costs, such as repairs and maintenance and kitchen supplies, which have grown faster than sales in recent years. We have invested heavily in training our field leaders recently to effectively manage the business culture in the restaurants they oversee.
Each field leader spent two full days in what we have called a business boot camp, and they have been training their GMs and reinforcing how to create an excellent guest experience while effectively managing their P&L. We also expect promo costs, which are another operating cost, will normalize in 2017 as we reduce reliance on giveaways and discounts. Marketing will either normalize or will remain at elevated levels in support of efforts such as TV advertising, which we would expect will result in higher sales and great leverage on the rest of the P&L. In addition, we have an internal goal of saving at least $100 million in recurring annual costs. This includes a review of everything we buy, both in the restaurants and in our offices.
It also includes capital expenditures, including the new, more efficient restaurant design Mark mentioned, which will reduce our investment costs from over $800,000 to around $760,000 in 2017, a savings of nearly $10 million in 2017 alone. As I mentioned previously, we plan to keep our underlying G&A costs before stock comp flat from 2016 to 2017. In dollar terms, that would result in G&A before stock comp of around $220 million in 2017. We chose not to fill about 100 staff positions this year, which were vacated via turnover and not refilled, or were approved in 2015, but were unfilled headcount adds, all of which will save around $10 million annually and will help our underlying G&A dollars remain at the current level.
Our teams are effectively supporting our restaurants by stretching the ratios a bit and focusing only on those activities that maintain safer food, lead to a better guest experience, that grow sales, or strengthen our economic model. We increased our field support significantly over the past three years and have nearly the lowest ratio of stores per field support staff than we've had in 10 years. We currently have nearly 300 field leaders, which includes area managers, team leaders, and apprentice team leads. Even as we build 200 new restaurants, on average, the field leader ratio will increase by less than one restaurant per field leader in 2017. Our restaurant teams and support teams are responding well to the challenge of more effectively managing costs and are energized by helping us restore our business.
As we challenged our restaurant managers at the AMC, one of our Restaurateurs, Philip Esposito, from our Newbury Park restaurant in California, reached out to me the very day after the conference with an idea to simplify some of our back office functions. His idea affects every restaurant and will save us more than $1 million across the entire company. I've seen similar commitments throughout the company, in our restaurants and in our offices, to be more disciplined and more discerning with every dollar we spend. Everyone feels ownership to do their part to help strengthen our economic model while we work to bring our customers back. We're confident that the approach we're taking will help restore our margin and our earnings capabilities without taking away from the customer experience.
In fact, the techniques our field leaders are teaching will result in scheduling and deploying the restaurant teams to deliver the best experience and the most delicious, safest food possible, which also leads to a better financial result. These efforts will allow us to generate respectable results in 2017, more importantly, we believe we will ultimately emerge from this challenge stronger than ever, as our cost structure will be leaner, we will be more focused than ever, our skill level throughout the company will be sharper, our economic model, with sales fully restored, will be stronger than ever. Thanks for your time today, we'll now open the lines for questions.
Thank you. If you'd like to ask a question, please signal by pressing *1 on your telephone keypad. If you're using a speakerphone, please make sure your mute function is turned off to allow your signal to reach our equipment. Again, that's *1 to ask a question. We'll go to our first question from David Tarantino with Robert W. Baird.
Hi. Good afternoon. Steve, I just have a general question on the overall operating approach going forward. I guess one of the hallmarks of the Chipotle business is how simple you've kept the operations and how focused you've been on delivering high quality. Now we're hearing that you're going to shift the focus a bit towards menu innovation and perhaps traditional marketing tactics or strategies. Just wondering why you think that shift is needed now versus perhaps just going back to basics on focusing on delivering the high-quality experience.
Well, David, I think that we're going to do both. We're going to focus on fundamentals and try to continually deliver a better experience through that method. We also want to try new things, again, to get some customers who may have lapsed, to reinvigorate our regular customers, to encourage them to come more, and perhaps to even entice new customers who haven't tried Chipotle. It's important that when you add something to the restaurant, you take something away. Although with the addition of new menu items, we're not going to take away other menu items, there are things that we can do to create efficiencies in the way we prep, in the way we get ready for business
In the way we find leverage. Finding leverage is interesting because there have been suggestions that we add breakfast, for instance. Well, I'm not saying that we won't add breakfast, but when trying to find leverage, our second make line now has an enormous opportunity. The second make line can attract or provide for catering, for delivery, for the mobile app to enable customers that are in line to get out of line and go to a tablet, and probably things that we haven't thought of yet. It's extraordinary leverage because the labor to produce the food is much less at the second make line because of the new technology. My point is that you need to balance everything in the restaurant, and so we will make some things more efficient, so we will be able to add things like a dessert item or a new menu item.
Again, if you look at some other examples of fast food places getting too complicated, I don't think they kept this idea of taking away whenever you add something. On balance, you still have the same amount of ability to focus on delivering the extraordinary experience that we've been accustomed to.
I guess just a quick follow-up to that, if I may. I guess you'd gotten the $2.5 million without any of those new menu items. I guess the question is, why do you think you now need new menu items to get back to those kind of levels versus just going back to the old operating approach? I guess I'm a little confused by that point.
Well, David, this is Mark. The new menu items, I wouldn't worry that we're going to slip into the traditional fast food model of introducing multiple new menu items wrapped with advertising multiple times a year. Our business really doesn't support that anyway, just because it's really one price for everything you get at Chipotle, unless you get something extra like meat or guacamole. However, there is a considerable amount of value to being able to advertise to people or to market to people something new. We've seen it now with chorizo. It's been very effective, which is why I mentioned in my prepared remarks that we're going to increase the amount of reach of that campaign four times.
What it does is it gives people a reason to reconsider Chipotle or just to add it to their routine, where they may have dropped on occasion. We're seeing it to be very effective. I wouldn't worry that you're going to see us switching to this model of repetitive menu additions with lots and lots of advertising around them. There is a real value that we've seen now very tangibly with chorizo in terms of adding something new every now and then. I really wouldn't worry that it's going to be a complete change in strategy at all.
Great. Thank you.
If you find that your question has been answered, you may remove yourself from the queue by pressing the star key followed by the digit two. Again, it's star one to ask a question. We'll go next to John Glass with Morgan Stanley.
Thanks very much. You highlighted a number of technology initiatives that will eventually drive sales. At least one of the competitors in the marketplace has done that. It's taken a toll on earnings for several years. It turns out there's a lot of training costs, there's technology costs. How much of that, if any, of those costs have you factored into 2017? Do you have a sense of not just the buying of the technology, but what the training costs might be for such initiatives?
Well, let me start, John, by talking to you about how the primary component of this thing works with regard to the technology. The second make line that Steve referred to, we've had one in our restaurants for many years, the classic version. This new version is much more efficient, whether or not you have the tech component added to it or not. It's just a better layout. It's got bigger bins. It's set up to be more efficient. That new line actually is slightly cheaper than our existing second make line. That line, that is without all the technology component on it, is actually a little bit cheaper than our existing line. We're going to move quickly to have all of our restaurants outfitted with that new line starting in March of 2017. All new restaurants will have it.
We're going to selectively retrofit high volume restaurants with it over the coming year. I think it's going to be relatively quick. Now, what I just described was the line itself without the technology component. It's completely set up to handle it. We add that line to the restaurants. As we see fit, we can add the technology component to it. It's not a tremendously expensive addition as well. You're not going to see lots and lots of the need to build big kiosks in the front of the house, that sort of thing, because this system actually, whether you're using it from a tablet inside the restaurant or from your phone in your pocket, uses a very similar ordering system to the one I described in my call, the new online ordering site that we've developed.
There's very little overall investment in this thing, and I believe that we can turn it around fairly efficiently as we roll it into our restaurants.
I would also add that often with the introduction of technology, there is a lot of expense in training to this new technology. With the second make line, this technology actually is very intuitive. It's actually easier to train a crew person how to operate the second make line with this new indicating system than with the old method.
In your plan for 2017, have you factored in costs, training costs, labor costs to do this, or is that not in the plan for 2017, you don't think it will be an impact?
John, we don't think it's going to be an impact. We think the labor is going to be neutral. We think over time it's going to be a labor saving. The labor per sales dollars on the second make line is more efficient than the labor per sales dollars on the front make line. Even though this is early on, what the teams have shown so far with this new approach is, they can produce food a lot faster, more accurately, and so there's really built-in efficiencies. If we didn't have second make lines, I would say this would be a major undertaking. It'd be a lot of training, and it would take quite some time.
The fact that virtually all of our restaurants already have a second make line, what they're doing right now, they're struggling to make the food on the second make line with a little piece of paper, with a lot of an abbreviation. What they're going to get instead of that, is they're going to have a very clear indicator, with display screens. They're going to be doing the same things they always have. With the restaurant that was retrofitted in N.Y., the team just intuitively stopped, just gravitated towards the display screens and, just intuitively just kind of made the food the way they always would, except they weren't stopping in between every ingredient or two to look at this little bitty piece of paper that's on the size of a register receipt.
It's early, but so far we don't think there's going to be an increase in labor. The equipment cost is not that expensive. There will be an impact on depreciation depending on how many we can roll out, but it's not going to be significant and not something that's going to affect our ability to get to the $10 EPS.
Got it. Thank you.
We'll take our next question from Jason West with Credit Suisse.
Thanks. Just a couple sales questions. I think you guys mentioned, Jack, you are running it down about 20% in October, but I also thought I heard you say that you have seen a full return to your kind of most frequent users, which I think is new. Could you try to explain what is going on there with the different customer buckets? You are still down 20%, but sounds like things are improving. Just trying to reconcile those issues. The other question on the comp outlook for next year, does that include any incremental pricing? Thanks.
Chiptopia was designed to go after our most loyal customers. It is a relatively small customer group, but it is a group that comes very often. We have sliced and diced our customer group a number of different ways, and when we look at our customer group, for those customers that come at least a couple times a month to one time per week or greater as one group. Another group would be customers that come, say, once a month. Another group would come two to three times a month, and another group would come once for every six months. When you take that most frequent group. By the way, all of those groups were negatively affected by anywhere between 20% and 25%, meaning they all dropped down their visit frequency.
In case of the infrequent customer, we just lost customers, and we lost them to the tune of about 20% to 25%. After Chiptopia, we recovered to a point of where our most frequent customers were back to 95% of where they were before. Some infrequent customers did not increase their frequency, but we were able to see some customers that were not coming as often as, let's say, once per week, now are coming once per week. We do not have the exact same customers we had before. One thing we have learned as we have gone through and we have gathered this data, mined this data that Mark talked about, we found that we have kind of a constant movement of customers that are coming. Some are coming more frequently, some are coming less. What Chiptopia did was encourage those customers either were coming frequently and reduced.
A lot of them did come back, and we encouraged some that were coming a couple times per month to come at least once per week or more. If you look at that most loyal customer group that are coming multiple times per month, that group, we have recovered 95%. The rest of the group is a much, much, much larger group of customers that come much less frequent. That is why the things that Mark talked about make so much sense. That is why TV advertising makes sense. That is why advertising things like chorizo, which is new news, makes a lot of sense. Dessert, once we decide which dessert we want to go with to try to go beyond a test market. These are things that you can advertise and appeal to a broad group of customers.
That's what our approach is to get that much, much larger, tens of millions of customers, to come to Chipotle until we can recover that group as well. In terms of the comp, we don't have any menu price built into the comp at all.
Great. Thank you.
We'll go next to John Ivankoe with J.P. Morgan.
Hi, thank you. Two, if I may. First on that $100 million of recurring cost savings, I think you called out $10 million of CapEx. Does that mean there is $90 million of OpEx savings that you expect in fiscal 2017 over fiscal 2016? Just as a clarification.
Not necessarily. There may be other things that we buy. For example, small wares. Most small wares would be OpEx. There might be some equipment that we'll do a more efficient job of buying and we might be replacing. It might not be a new store impact, but we'd be replacing equipment, let's say grills, for example. We're looking for a reduction in expenses or in what we pay for everything. There may be some additional CapEx, but I would say the majority of the $100 million is going to be in operating expenses.
Okay. Secondly, the Chiptopia program, obviously, it was successful with your most frequent customers. Given some of the complexity regarding the program, maybe some of your not-as-frequent customers didn't know necessarily how to use it or what their benefit would be or, going that number of times in a month would be a lot to ask of someone that doesn't go to the brand, for example, once a week. Is there a way to approach loyalty in a different way that would have a broader appeal? Is that potentially part of the plan in 2017 and 2018 to have something that's simpler and that will be a permanent part of your offering?
Yeah, John. An interesting thing, which I mentioned in my remarks, I think that's worth noting, is that over the last six months, we saw 30 million customers that were either new or people that we hadn't seen before, come into Chipotle. This is thanks to this new data capability that we have. Additionally, we're now able to reach a very large number of those people. We have dramatically improved our ability to reach out to these customers. They're in all frequency categories, by the way. This 30 million people range from one to more than 10 times in terms of number of visits during that six-month period. They're all across the different frequency bands. We have such an extraordinary new ability to reach them, that we're evaluating exactly how best to use that in order to create incentives for them.
We don't want to rely on a continuing basis on discounting. Obviously, that cheapens the brand and creates a sense of entitlement, ultimately, such that people just begin to expect it. We're very carefully evaluating how that's going to look. To answer your question, I would say, yes, we're going to look at how we can use this new information we have about our customers and this really amazing ability to track whether or not what we do with them drives them in. We can determine now with a very high degree of fidelity, whether or not an offer or an ad of a particular type or any other type of promotion or communication actually resulted in a visit from these customers.
I don't want to commit right now to exactly what we're going to do, because it's really just been over the last few months that we've developed this capability. We're absolutely going to use it in a way that allows us to create ongoing incentives to create the momentum for these less frequent customers to become more frequent. I'm very excited about just the simple idea that so many new customers came to Chipotle. It says a couple of things to me. First it says, even though it's tempting to think that, well, people are still reluctant to come to Chipotle, with that volume of people coming in as new customers, I think that's unlikely. That's showing a pretty major shift in people's attitude. Also, those somewhat infrequent new customers have extraordinary potential to become frequent.
They're the ones who have the potential to create great comps for us in the future. We're going to be doing whatever we can to make their stay at Chipotle as enjoyable as possible. I can't give you a firm answer, but I hope that helps you understand how we're going to be looking at it.
Thank you.
We'll go for our next question to David Palmer with RBC Capital Markets.
Thanks, and good evening. First, a follow-up on some of that questioning about the marketing. You've mentioned a few things, the new product news, national TV, all these things will be part of Chipotle's future, potentially, and you've tinkered with loyalty and now looking to make a bigger push behind mobile order and maybe even delivery. Based on the analysis you've done so far, because it's hard for us to assess, what potential changes to marketing among these and maybe some other stuff do you believe will be the most promising and most likely that we will see?
Well, they all have promise in their own right. I think, though, that as I mentioned in answering the last question, we've seen very large numbers of new customers experiencing Chipotle. I think there's extraordinary potential for us to continue bringing in large groups of new customers, and that's the reason why we're currently testing television. Television isn't a panacea. It's going to be a component of larger campaigns that we do, but it's very efficient when we get to the level of doing it nationally. I believe that that probably has some of the most significant potential for us to bring in lots and lots of new customers. Assuming that we create a great experience for them when they come in the restaurant, they will then turn into more and more frequent customers.
I suspect that amongst those things, that has huge potential, but I will follow that with a very close second, which would be the increase in our digital promotion and the capability of online ordering. I just want to take a second to tell you a little bit more about what I'm talking about with regard, for example, to Smarter Pickup Times, which is the ability for somebody to come into Chipotle and get an order very quickly after they ordered it. We've been conducting a test over the last several months of this technology, and through our analysis of that, we found that when people place orders at Chipotle, there's a very significant lag between when they place the order and when our restaurants allow them to come pick it up.
This new technology actually eliminates our restaurant's ability to set that time gap, and instead sets it automatically based on the load of the restaurant. We're seeing the potential to do as many as one entrée per minute using this new system. The marketing that goes around that encourages people to place their orders via digital, assuming then again, those orders are fulfilled in this efficient and excellent way, I think has a tremendous amount of potential. I'd put that as a close second to the top of mind marketing that we're going to be doing.
Thank you.
We'll go next to Karen Holthouse with Goldman Sachs.
Hi. Thanks for taking the question. I guess just thinking about some of the other pieces or part of improving the digital experience at the store. I know the POS system has been a part of upgrades were needed to be able to move forward. Is everything else in place that you think a lot of this can be pretty plug and play once it's in store? Are there technological or technical hurdles you think you're still going to be working your way through?
I think that most everything is in place, and I say that wanting you to know that a couple of the key pieces are actually going to be rolled out throughout the month of November. We have, as you mentioned, we've had some issues with regard to payments in particular, and the new responsive online ordering website that we've talked about in this call and some other, including catering, will now have a more streamlined payment process. I won't go into the technical details as to why it's more efficient both for us and for the customer. Having gotten over that hurdle actually enables us to do a lot of the things we talked about without further technological investment.
I do want to caution, though, that of course, when you're adding anything that consumers are interacting with and new technology for customers, there's always room for things to go wrong, and we're going to need to iterate and refine those things. I don't see any large technological hurdles that we need to overcome in order to implement this. We've overcome a lot of those over the last 18 months, actually, and we're just now in November going to be taking advantage of a lot of them.
Just a follow-up on the digital, the more one-to-one marketing. Is the data set that you're working off of? Because my impression or my understanding is that the average consumer isn't scanning or swiping a dedicated Chipotle card at Chipotle. Is that working off of last names and last 4 digits of credit cards? How does that evolve or change as we move towards Chip and PIN implementation when you don't get that data anymore?
Well, yeah. We are using a combination of last name and 4 digits of their card, also with a unique combination of both Chipotle-owned and commercially available databases to create actual profiles of each customer. We have the ability to now, as I mentioned, reach many of our customers using this technology. Chip and PIN is something that we have yet to implement at Chipotle for throughput reasons, and we very much hope that we're going to leapfrog Chip and PIN. I feel reluctant to go too far into the technological reasoning behind that without Curt, our CIO, weighing in on it.
There are more efficient ways of making payments than Chip and PIN, and it doesn't seem to be, even though it is required by the credit card companies, it does not seem to be the technology that's going to be readily adopted by choice for consumers. We are simultaneously working on a variety of payment options ourselves, and we expect that those will augment any loss should we ever implement Chip and PIN.
Great. Thank you.
We'll take our next question from Jeff Farmer with Wells Fargo.
Great. Thanks. I'm just curious, what are the pros and cons of maintaining almost 10% unit growth rate in 2017 on the heels of what looks to be like almost 12 months of a 20%+ average transaction decline? Just curious why you guys are sticking with the big numbers instead of potentially taking it down 10% or 20% just until the transactions stabilize a little bit.
Well, Jeff, we took a careful look at our whole portfolio, and we did pare it down, and we pared it down where we eliminated either remote sites that are hard to operate or newer developing markets where the risk of low sales and the risk of trying to staff a team and attract customers in an area that doesn't have perhaps a strong brand awareness or brand appreciation. We pared those down. To cut the openings even further, there just isn't an advantage to doing so. The impact when we open up restaurants on our existing comp is minimal. It's always been less than 1%. It's still less than 1% this year. It's not like there would be a significant comp rebound. We have the pipeline. Other than paring down, as I mentioned, we have a strong pipeline.
There just isn't an advantage. The recovery doesn't get jump-started just because we would peel off another 25, 50 or 100 openings. There wasn't any advantage to doing so. We thought we did the responsible thing in doing what we could to shore up the quality of the portfolio. Going beyond that just didn't offer any benefit from what we could see.
Okay, that's helpful. Just one follow-up. On the potential timeline for the national TV and heightened digital capability rollout, you guys mentioned it a couple of times, theoretically, how quickly could we see TV and under what circumstances would we see TV?
Well, right now TV is being tested in three markets between October 21st and November 20th. We'll evaluate that on a number of different levels, that television in two of those three markets is running concurrently with other supporting advertising, which is normally the way we would do that. Television doesn't stand on its own, typically. As a result of the test, we will integrate television, as I mentioned, probably at a national level. That's where the great efficiencies come in when you're making point buys like that. We would integrate that into the most likely the campaign that we have launching in spring. Now, that doesn't mean that we couldn't add some selective TV for certain reasons before that. I would say that certainly by March 2017, you'd see, if the test is successful, television integrated into the campaign.
All right. Thank you.
We'll go now to Joseph Buckley with Bank of America.
Thank you. Two questions. One, curious what the second make line today is contributing to sales. Are the sales off that second make line a significant part of the business today?
Yes.
About 6%, Joe.
Okay. Bigger picture question. As you shift the focus next year to a little bit more on the unit level economics, are you coming to accept that the base level of average unit volumes, the base level of restaurant margins may be lower and resetting that as the bar from which you'll grow going forward? Is it somewhat of a concession that recovering the full sales level seems less likely?
No, we're not conceding at all, Joe, that our top-line sales can't be dramatically improved over the coming years. We are just recognizing that in the meantime, though, it behooves us to do everything we can to be as efficient as we can be so as to drive the very best margin possible, even during a time where our volumes are still negatively affected. Like I said in my remarks, we're not going to drive those costs down to the point where we're negatively affecting the guest experience or the ability that our field teams have in order to continue to create really great restaurant teams and have great restaurant operations.
Okay. Thank you.
We'll take our final questions from Nicole Miller with Piper Jaffray.
Thank you. Good afternoon. What level of share repurchase is accounted for in fiscal 2017 earnings guidance? What will the cadence of new store development look like?
Nicole, I would expect a similar buy as what we've seen in the last one or two months. We bought about $1 billion worth aggressively. That took us through about April or May or so. I would expect us to be at more of this kind of steady level. We'll want to keep a strong balance sheet somewhere in the half billion dollars or a little higher on the balance sheet. Then we'll take excess cash flow and use that. If in your model, if you take excess cash flow and assume that we kind of keep what's on the balance sheet or about what's on the balance sheet today and use any excess to buy back stock, I think you'll be in the ballpark of what we'll actually do.
Now, we'll be opportunistic, and we'll be more aggressive if the stock drops and less aggressive when the stock increases. Generally, that's the approach we'll take.
To answer the question with regard to the new restaurant openings, we've said that it'd be between 195 and 210 restaurants for 2017.
Yeah, the cadence will be fairly level loaded throughout the year, Nicole, if that's what you're asking.
They tend to pile up a little bit in Q4 typically, we'll level load as best we can.
Thank you.
Thank you, Nicole.
Thanks, Nicole.
That concludes today's question and answer session and brings to the end our conference. We thank you for your participation. You may now disconnect.