Good afternoon, and welcome to the Bloom Energy fourth quarter 2018 earnings call. At this time, all participants are in a listen-only mode. Later, we will conduct a question- and- answer session, and instructions will follow at that time. As a reminder, this conference call is being recorded. I would now like to turn the conference over to Mark Mesler, Vice President of Finance and Investor Relations at Bloom Energy. Please go ahead.
Thank you. Good morning, all, and thank you for joining us on Bloom Energy's fourth quarter 2018 earnings conference call. To supplement this conference call, we have posted to our investor relations website our Q4 2018 shareholder letter, as well as supplemental financial information that we will periodically reference throughout this call. Please note that this call contains forward-looking information regarding future events and the future financial performance of the company. We caution you that such statements are predictions based on management's current expectations or beliefs. Actual results may differ materially as a result of risks and uncertainties that pertain to our business. We refer you to the company's SEC filings, including the company's quarterly report on Form 10-Q for the fiscal quarter ended September 30th, 2018. These documents discuss important factors that could cause actual results to differ materially from those obtained in the company's projections or forward-looking statements.
We assume no obligation to revise any forward-looking statements made on today's call. During this call, in our Q4 2018 shareholder letter, we refer to GAAP and non-GAAP financial measures. These non-GAAP financial measures are not prepared in accordance with generally accepted accounting principles. A reconciliation between GAAP and non-GAAP is included as part of our Q4 2018 shareholder letter. Joining me on the call today are KR Sridhar, Principal Co-Founder and Chief Executive Officer for Bloom; Randy Furr, Chief Financial Officer; and Matt Ross, Chief Marketing Officer. KR and Randy will review the operating and financial highlights of the quarter, and then we will take questions. I will now turn the call over to KR.
Hello, this is KR, and good afternoon to all of you, and welcome to our Q4 2018 conference call. I'm very excited to share with you my perspective on our Q4 and FY 2018 results, capping our first fiscal year as a public company. As we enter 2019, we have a strong foundation in the fundamentals of our business. One, a strong and increasingly diversified sales backlog reflecting the mainstream appeal of our offering. Two, continued cost reduction. Three, a unique, innovative, and proven technology platform in the Bloom Energy Server. As you know, we provide three key benefits that our customers highly value today. One, lower cost of power; two, lower emissions; and three, higher resiliency, including the option for uninterruptible, always-on power. It is clear that this value proposition is relevant and important to an expanding customer base.
I'll focus on four key areas with you today: growth, quality and diversification of our backlog, cost reduction, and outlook for 2019. Let's start with growth. We grew 2018 acceptances by 30%, or 187 systems in comparison to 2017. In 2018, we had 809 acceptances compared to 622 in 2017. In Q4 2018, we grew acceptances by 28%, or 56 systems over Q4 2017. We had a record 257 acceptances in Q4 2018 compared to 201 for Q4 2017. We are very happy with this growth. When we look ahead to 2019, key metrics to consider are the product and installed backlog volume and quality of that backlog. I'm pleased to tell you that we exited 2018 with a strong product and installed backlog of 1,384 systems, 70% higher volume than our actual customer acceptances in 2018.
Based on this product and installed backlog, plus anticipated bookings that we expect will convert to acceptances this year, we expect to have year-over-year total percentage revenue growth in 2019 in the 20s. Let's look at the quality of our backlog. Diversity of sectors, geographies, and the potential for future growth in these areas are key metrics to evaluate the quality of product and installed backlog. We have executed on a deliberate strategy to diversify our growth opportunities as our value proposition becomes more mainstream, both geographically and by industry sector. Data tells the story. In 2011, 100% of our acceptances were in California. In 2017, that number was still 87%. In 2018, it was 42%.
Our backlog exiting 2018 suggests that our acceptances in California will be less than 30% of our total 2019 acceptances and not more than 15% in any single utility service territory in the state. We view this diversification to be very important to our business. When we look outside of California, our backlog shows good geographic diversity in six Northeastern states, Japan, and South Korea. Like many companies you are familiar with, Bloom Energy's business today has seasonality to it. If you consider years 2016, 2017, and 2018, our average acceptances in Q4 cumulatively were higher by over 50% than our cumulative acceptances for the Q1 period in the same years. In general, second half acceptances are significantly higher than first half acceptances for Bloom Energy.
We expect 2019 will follow the same cadence, and in fact, will be impacted even more by seasonality as our product and install backlog expands in the Northeast, where we face winter weather challenges such as the recent polar vortex, as opposed to the milder, construction-friendly winters in California. Our business in Korea will help to mitigate this in Q1 2019. When we look at our current bookings, new orders span a highly diversified set of industries, including data center, cloud services, healthcare, retail, hospitality, advanced manufacturing, higher education, real estate, government, and utilities. This diversification mitigates our exposure to any one location, any one utility service area, or market sector. We have intentionally built this diversity into our product and install backlog, and it enables us to sustain ups and downs in any one area of the economy or any one geographic area.
Let me take a few moments to talk about the momentum we have in the Korea market. We see Korea as a very attractive market to grow our business now and in the future. We achieved our first order in South Korea in December 2017. That 8.35 MW project with KOEN is now operational less than a year after the PO. In November 2018, we announced a strategic agreement with SK, the third largest company in South Korea, to be our preferred distributor of Bloom Energy Solutions in Korea. In December 2018, with SK, we won four utility scale projects with customers including KT Corporation and Korea Midland Power, a KEPCO power generation company. We plan to fulfill these orders in Q1 2019 to establish our brand and reputation in a strategic growth market, as well as mitigate weather-related risks in the Northeast in Q1.
We see it as a worthwhile and prudent investment. These acceptances will be below the margin targets for the rest of our acceptances. Additionally, we will have non-recurring expenses relating to certification costs to operate in Korea and the establishment of an infrastructure for operating and maintaining our energy servers in Korea. The non-recurring cost impact and lower ASPs will be reflected in our margin guidance for Q1. We consider this an investment that will pay off as early as the second half of this year, when we expect our margins in Korea to be in line with our corporate targets. To further fuel our growth, it is the right time for us to be investing in sales and marketing. Historically, our investment in sales and marketing has been modest, well below typical investments in this area for a disruptive innovation company like Bloom.
We are now carefully investing in the areas that will help us to enter new markets and expand our penetration in existing markets. Let us turn to cost reduction. While we are continuing to reduce cost in our Bloom 5.0 commercial platform, as discussed previously, we have ramped up R&D on our new Bloom 7.5 platform. We are on track with our plans on this very important program. We are ramping up our R&D in 2019 to build Bloom 7.5 prototype units. Bloom 7.5 should offer our customers more power in the same footprint and with a further increase in our already world-leading fuel efficiency. This platform will also enable us to continue to drive product and service costs down significantly. Consider this. Bloom 7.5 should deliver 50% more power in the same physical footprint as our current generation, bringing about a step change in reducing our product cost.
This, combined with continued manufacturing process innovation and volume growth, give us very high confidence about continuing our rapid cost-down trajectory. While we are investing in our technology to drive cost reduction, we are also investing in R&D to broaden our offerings for the future. An example is our investment in biogas initiatives, enabling us to provide expanded net zero carbon solutions. For our customers, biogas fuels AlwaysON Bloom solutions deliver uninterruptible 24/7 power, enabling them to both lower their emissions and to power through events that otherwise would pose a risk of operational disruption. We believe now is the time to press our innovation advantage and to lengthen our lead in power solutions that combine low or no emissions with always-on resilience. For this reason, we are also investing in a range of always-on solutions tuned to the needs of different industries and applications.
We hope to announce some exciting customer wins in 2019. Let's summarize what all this means for 2019. Based on our product and install backlog, I expect to see revenue growth in the mid-20 percentage points year-over-year. I also expect our gross margins to be in the mid-20s in the second half of 2019. Together, these expectations should put us well on our way to our long-term business model of 30% revenue growth and 30% gross margins. Our actions and investments are geared to provide not only healthy growth and margin, but also a diversified portfolio of opportunities that has the potential to make our business more predictable and less susceptible to localized market and geographic disruptions. With that perspective, I would like to invite Randy Furr, our Chief Financial Officer, to walk you through our operating results for Q4 and estimates for Q1 2019.
Over to you, Randy.
Thanks, KR. Throughout my prepared comments, I'll be referring to the slides in the earnings call presentation that Mark referred to earlier. First, some highlights. Note that all profit numbers that I reference will exclude stock-based compensation. On to slide three. In summary, a very respectable quarter. Acceptances were 257 systems, a record. Revenue was $213.6 million, up sequentially by 12.3%. Non-GAAP gross margin come in at 18.1%. Our non-GAAP operating income was $4.7 million, with adjusted EBITDA coming in at $14 million. Adjusted EPS was a loss of $0.12. Year-end product and installed backlog rose almost 17% to 1,384 systems, with an approximate dollar value of $773 million. This backlog excludes any long-term service agreements, as we recognize those billings as occurred on an annual basis, and it excludes future electricity revenue as well. As a reminder, service and electricity revenue in 2018 totaled $138 million.
We ended the quarter with $348 million in cash and short-term investments, and this excludes $36.4 million of PPA cash. In addition, given the December quarter was the fourth quarter of our fiscal year, I wanted to share some annual highlights as well. Acceptances were 809 systems, a record as well, and up 30% from 2017's 622 systems. Revenue totaled $742 million for the year, and this includes the one-time top-line benefit of $45.5 million related to the federal ITC reinstatement, as this was for 2017 ITC benefit that was received in Q1 of 2018. Backing out that one-time benefit, the actual 2018 revenue was $697 million, a number that was up significantly from 2017's $376 million, where we did not have the federal ITC benefit. Non-GAAP gross margin come in at 21.4%. Non-GAAP operating income was $26.4 million, and adjusted EBITDA was $66.9 million.
During our call last quarter, I walked you through our business model. I explained how we define the term system here at Bloom, how an order gets into backlog, what is meant by an acceptance, and how our reliance upon third parties can impact the timing of those acceptances, how the federal ITC impacts Bloom, and how you need to factor in ratable versus upfront revenue recognition when looking at historical financials. As such, I'm not going to dive into those topics today, and we'll move directly to providing some color on our Q4 results. As I mentioned last quarter, we will only disclose our product and installation backlog once a year at year-end. Once again, orders generally come in lumpy, but reporting on an annual basis helps us normalize that lumpiness.
We ended the year with 1,384 systems in product and installed backlog, a 16.6% increase of the 12/31/2017 number, and as KR mentioned in his prepared remarks, a backlog number that is over 70% higher than the 809 systems recognized as revenue in 2018. Our backlog goal is to maintain 9- 12 months of product and install backlog at any point in time, and we're at the higher end of that window as we enter 2019. Referring to slide four, the 257 acceptances translated to $213.6 million in revenue, up 12.3% from Q3's $190.2 million, and up significantly from last year's GAAP revenue of $123.3 million. The quarter-over-quarter increase in revenue is attributable to the growth in acceptances, which includes the Korean orders.
Recall that because we are not responsible for the installation of these systems, our revenue recognition here is based on when the systems are received at port and not when they're generating electricity like our U.S. C&I business, where we generally are responsible for the installation. On a year-over-year basis, from a GAAP perspective, the majority of the increase was attributable to the realization of the federal investment tax credit, which was not available in 2017. Note from an operating metrics perspective, and when measured against billings, year-over-year revenue growth for the quarter was 65.2%, with some of the increase due to the increase in acceptances, but the majority due to the reinstatement of ITC in 2018. On to slide five. Gross profit, excluding stock-based compensation, was down from $39.5 million in Q3 2018 to $38.7 million in Q4, a 2.1% sequential decrease.
This was in line with expectations as 58% of our Q4 acceptances were international, where ASPs are lower. Year-over-year comparisons are not particularly meaningful given the absence of ITC in 2017. As most of you know, we do provide some quarterly estimates, and in our Q3 shareholder letter, we provided you with Q4 average sale price estimates, as well as average total installed system cost estimates. For Q4 2018, both the ASPs and the TISC came in in line with those estimates. However, as I previously emphasized, the real key metric here is the delta between the two, which represents our margin on the equipment and installation of the acceptances during the quarter. The midpoint of the estimated ASP and TISC yielded a delta or margin estimate of $1,455 or $1,455 per kilowatt.
As you can see on slide five, our actual margin delta was $1,412, a number generally in line with the midpoint of our estimate. As you can see on slide six, non-GAAP operating income for Q4 was $4.7 million, excluding stock-based compensation. This is down $900,000 from Q3's $5.6 million, again reflecting the mix of international shipments for Q4. The $4.7 million operating profit is up considerably from Q4 2017's operating income loss, realized both from a non-GAAP and an operating metric standpoint. Again, primarily due to the increase in acceptances and the reinstatement of ITC in 2018. Our adjusted EBITDA come in at $14 million for the quarter. Non-operating expenses were per plan, adjusted EPS come in at a loss of $0.12. Turning to the balance sheet on slide seven, we ended the quarter with $384.8 million of cash and short-term investments.
This includes $36.4 million of PPA cash, excluding PPA cash, we ended with $348.4 million of cash in short-term investments. Free cash flow, which we define as cash flow from operations less capital expenditures, was a negative $55.6 million. This use of cash reflected the mix of international revenue for the quarter. The payment terms with our Korean partners specify 60 days after arrival at the shipping port. This contrasts our domestic revenue, where we generally receive progress payments with the full amount of the contract received no later than seven days after acceptance. This mix translated to approximately $57 million of Korean shipment-related accounts receivable that was billed in Q4 and will be received in Q1. If we factor in the Korean receivable, we would've been cash flow neutral for the quarter.
With the exception of the accounts receivable just discussed, our working capital metrics came in line with expectations. Referencing slide eight, days of sales was up eight days from Q3 to 27 days, again, reflecting the international shipment terms. Our days of inventory outstanding was down 14 days from Q3 to 73 days. This reflecting the record shipments and acceptances for the quarter. Our payable days was down from Q3 by one day to 33 days on normal business cycle variations. Changing the conversation to our outlook. In Q1, we expect acceptances to be between 215 and 245, ASPs to be between $6,750 and $7,050, with our total installed system cost to be between $5,600 and $5,900. I'd like to add some color to our outlook. First, I'll start with acceptances. This is in the neighborhood of 10 lower than what we presently see as consensus for Q1.
This simply reflects us being conservative given our past challenges in this area. Also included in our outlook are additional shipments to Korea. A little over 50% of the quarter's total shipments will be destined for Korea in Q1. This influences our ASPs a couple of ways. First, we have no installation revenue for Korean revenue, and that translates to a lower ASP as compared to U.S. revenue. Keep in mind that we also have no installation cost as well. Second, as mentioned during last quarter's call, international shipments and revenue today carry an overall lower ASP than does domestic, and this translates to a margin lower than our target. You might ask, why do any Korean volume today?
The reason is, later this year, Korea should provide us with larger volumes as well as margins that will enable us to hit or exceed our communicated targeted margin. In addition to the lower acceptance volume included in our outlook for Q1, we are taking a volume hit to our quarterly total product cost as production builds in Q1 will be less than in Q4. Less fixed cost absorption adding up to a higher quarter-over-quarter cost of goods sold as a percentage of revenue for the quarter. This, of course, will start to move favorable beginning in Q2. I also want to add that we expect operating expenses will be up approximately 12% quarter-over-quarter. This incremental spend is related to R&D as we ramp up our next generation Bloom Energy Server and invest in our always-on solution and our new biogas product.
Also contributing to the additional operating expenses are investments in our sales and marketing efforts. What all this adds up to is an expected non-GAAP operating loss for Q1. Let me summarize what's driving that loss. One is once again top line and mix related, as our outlook represents 27 fewer systems at the midpoint of our estimates, with a little over 50% of our revenue being Korea. Two is the negative impact on cost that comes from fixed cost absorption as a result of lower build volume. This impacts profitability in the $3 million range. Finally, you need to factor in our planned additional investment in R&D and sales and marketing. This resulting in about a $4 million impact relative to Q4 operating expense spending. I will close with some comments on the second half and full year for the 2019 year.
We have previously communicated our 30% growth and 30% non-GAAP gross margin longer term goals. We believe we are on a path to achieving both. In fact, we believe for the second half of this year, we will be solidly in the range of between 25% and 30% for both revenue growth and non-GAAP gross margin. Once again, thank you for your time, and now I'd like to turn the call back to the operator for Q&A.
At this time, I would like to remind everyone, in order to ask a question, press star, then number one on your telephone keypad. We'll pause for just a moment to compile the Q&A roster. Your first question comes from Paul Coster with JP Morgan. Your line is open.
Yes. Thank you very much for taking my questions. I've got two. You talked about higher margins on the South Korean product in the second half of the year. Can you just elaborate a little bit on why the ASPs and margins are going to improve?
Yeah, Paul. Good question. Look, it's fairly simple. We have an agreement. We know what our ASP is. That's not gonna fluctuate. We also have, through our detailed budgeting process, a pretty good idea of where our product cost is going to be later in the year. It's a simple difference between those two that gets us to where we're comfortable that's going to be in the communicated targeted margin range that we had before. Why is cost coming down to enable us to reach that is a lot to do with just our continued cost efforts down and the increased volume that we're going to have, our fixed cost spread over a larger volume that's going to give us that cost point that will let us achieve that targeted margin.
Okay. Got it. I've got a lot of other questions, but I'll confine myself to one other, and that is, no real commentary around the installation timelines and delays or anything of that nature. It suggests to me that maybe you're starting to get on top of that part of your business challenge. Can you just tell us how we're doing in terms of sort of delivering in a timely manner and eliminating some of the variables that were affecting the acceptances domestically?
Paul, this is KR. Yeah, there was a lot of speculation about given what we had done earlier, whether in Q4 we will meet our acceptance target and what was going on, especially in California and how that would impact us. You can see from the numbers that we put in a strategy, we have added both in terms of leadership as well as other important skill sets to that area and made installation timeline a key area of focus for the company. We are beginning to see the results. Stay tuned before the next earnings call we may give you some news and color on the additional leadership that we have brought on to Bloom to help in this area. You will see that it's already paying very good dividends. Randy, do you want to add?
I'll just add that we're trying to be conservative in our forecast and our outlook going forward, that's kind of reflected in the Q1 outlook as well.
Yeah, we appreciate that, Randy. Thank you.
Your next question comes from Pavel Molchanov with Raymond James. Your line is open.
Thanks for taking the question, guys. I know you're not explicitly guiding on TISC, you're indicating an uptick in that cost from Q4 levels. Is it fair to say that once the startup issues in Korea subside, the downward trend in your TISC should resume?
Yes. You hit a very good point, you're on target. I want to slightly word it a little bit different. It's not that we have any startup issues in Korea. It's just that there's obviously anytime you enter a new market, there's expenses associated with that. As we well know, Bloom has come up, invented the concept of the Power Tower in Korea. Obviously, we have a large number of systems in Korea today, and the Korean electrical standards are different than, say, the U.S. standards and making sure all of our products conform to those specifications and standards. There's some expense associated with that, and we're experiencing some of that expense. Things are progressing and moving very well in Korea, and we're very proud of those installations over there.
You're right, there are some incremental expenses that we're seeing showing up in our cost of goods sold, which is being reflected in TISC here that you will not see later in the year.
The key point that Randy made is, again, I want to emphasize, Pavel, it's not an issue. It is a real non-recurring expense in terms of a starting up in a new area, is how you should categorize that. The reason, if you're looking for a data point to give you confidence on there are no issues, you have to think about this. December of 2017 is when we had a purchase order to do a 8.3 MW Power Tower. By fall of the following year, in less than nine months, we had the system up, and by the end of the quarter, we got certification, and we were having that one tower operating, one of a kind, the very first time. There are no issues, but there are startup costs associated with it.
Okay. One more regarding Korea, if I may. To clarify, the distributorship that you announced today with SK D&D, that's a separate agreement versus another division of SK Group that you had announced last year. Is that right?
That is absolutely correct. Our first agreement that we announced was with SK E&C, and they do utility scale projects. They do very large projects for large corporations, things of that nature. SK D&D is an independent separate business unit still within the same conglomerate. SK D&D focuses on real estate and large buildings in Korea, where they are the leader today in on-site energy management and storage solutions, as well as solar solutions. They are super excited to bring a base load to that mix, thereby they can offer what will be truly microgrid solutions to a customer, where they can have always-on power in addition to benefiting from the lower carbon footprint.
This would be whatever opportunities come from this particular partnership will be additive to what we have already signed on the utility scale with SK E&C, so you can see why we are super excited about Korea as a market.
Okay. Thank you very much, guys.
Your next question comes from Stephen Byrd with Morgan Stanley. Your line is open.
Hi, good afternoon.
Good afternoon.
Go, Stephen.
I wanted to look at the first quarter guidance. I believe a majority of the sales will be in Korea. I guess from a positive point of view that to me seems to highlight potentially less execution risk in terms of other projects. I wondered if you could just talk at a high level in terms of the kinds of execution risk to achieve on the balance. I guess when you ship internationally, you are able to book those sales. What is the degree of execution risk in your first quarter? I know you've been asked in a variety of ways just to think about execution risk. What kinds of challenges would you have if weather's not optimal or other kinds of risks? How should we think about sort of execution risk for the first quarter?
Yeah, again, great question. I'm going to give a little bit of an answer here, so bear with me a second. As we sometimes talk about, the great thing about Bloom is we have great technology, we have great manufacturing capabilities, great quality, and really the lead time from the time we get an order to the time that we recognize revenue on most of the U.S. C&I business is very little time to manufacture the product and actually deliver them to the site and get them up and working and turned over. Two-thirds of the time or this 9- 12 months timeframe that we take really centers around the design, all the field work we need to do prior to that design and during the design, all the customer, landlord, utility, government approvals for the building permit.
All of that stuff usually takes six to nine months to accomplish. In all of our installations, each site, we have a detailed project schedule, and we go into that knowing a date. 100% of our backlog today has a date out there that we think we're going to get an acceptance because we do that plan. Clearly, we deal with so many third parties that there's risk in that, and we've highlighted the risk, I think, last quarter to do with the hurricanes or the natural disasters, the wildfires, how utility crews are pulled off and sent to regions, and these are the guys that maybe come out and do the tie-ins or do the inspections, and it just plays havoc with the calendars, right, and we have those issues. We don't want to miss our quarters. That's obvious. None of us want to do that.
We feel bad. That's not our goal. How do we deal with that? We deal with that by having a pool of systems. Every quarter we come up with a pool. This quarter we had 257 acceptances there, and I can tell you the pool was much larger than 257. All we have to do is make sure that we yield a number out of that pool that lets us meet the expectations that are out there. How we deal with that is just as I said, we try to have plenty of backups. If something was to go wrong, these things that go wrong that impact the timing of the acceptance are really to do with inspections from third parties.
Should something happen there that's not in accordance with our schedule and cause the project to be delayed, we have this other pool of systems that we can go to and rely upon to be able to make the quarter. What we're doing, certainly after last quarter, is we're trying to be more conservative there with the estimates that are set or the outlook for the quarter based upon the size of the pool we have. We're trying to get better at execution and planning. With that said, it's inherent part of the business we have and the way we're going to deal with it is having this pool.
Stephen, let me add to what Randy said. He covered most of the points. In my opening remarks, when I made the comment of the importance of diversity geographically, a key underlying statement that I should have made probably, and I will make it now, is when you look at it from an installations perspective, particularly. What do I mean by this is, take Q1 as an example. I walked you through how, as we grow and open up new markets, our reliance on California in terms of the total acceptance has been diminishing all the way from 100% to less than 30% this year. That's a good thing, but also understand the other six states that we operate are in the Northeast.
The East Coast, as we look at January, if January is any indication, we don't know what the groundhog is going to do, or whether February and March are going to be January or not. We have two choices here. We can go to the international market and fulfill this, or we can pay enormously large expedite charges to catch up for any weather-related issues and still try to make our quarter in the East Coast regions. As long as our customer is willing to wait, there's no reason to go through those additional expenses which will eat into the margin, and you can swap things out back and forth. Having this flexibility of geography, multiple zones, is really helpful. I think to Paul's question, we answered another question. We are doing two things simultaneously.
One is trying to be really, really good even in the face of contingency, of managing all these risks, making our execution really good within our installation group by adding skill sets, adding to the leadership, doing all that. In addition to that, when we have optionality like what we have of international versus East Coast versus West Coast, the combination of all these things is what ultimately will allow us to execute, because the uncertainties of a forest fire or somebody behind a desk in a city government in the last minute denying you a permit or a utility person not showing up to give you a connection is something we cannot eliminate. By using these strategies, we can still deliver on what we need to do, and that's what we're focused on. We demonstrated that in Q4, since you asked.
The forest fires did not hinder us from meeting the numbers we put out for you in Q4. We're hoping to do the same thing in Q1.
It's extremely helpful color . Thank you very much. My next question is fairly high level. Beyond Korea, where you've had very clear success, would you mind just talking a little bit more internationally at where you're seeing the opportunity? You don't need to be too specific, I'm just interested in what kind of international opportunities you're seeing. What are you most excited about over time? Not necessarily in any particular quarter, just at a high level, what are you most
Stephen, Matt's in the room, for those of you on the call, Matt Ross is our Chief Marketing Officer, I'll have him answer this question.
Thanks, KR. Hi, Stephen. Keep in mind, I think we made the statement earlier about very low investment in sales and marketing. We pick our shots, we're very careful about doing that. We don't have the coverage to go broad. We pick our shots very carefully. As you can see that Korea, after being in that market for one year, is already scaling very rapidly. Another market that we see some real potential in for 2019 is Japan, that's because of the development of the Power Tower that gives us a very effective energy-dense platform for deploying in Japan. It opens up some of the kinds of opportunities that we've seen grow very rapidly in the U.S. For example, edge data centers that are in dense urban areas. Well, we can do that with the Power Tower now.
The economics for Japan are looking pretty favorable this year, taking into account our general cost down as well. Another area is India, where we are working on biogas projects. We think those will be small initially in scale, they're strategically very important. We see an opportunity in India, probably not very large numbers in the next couple of quarters. Strategically, again, very important. Those are a couple of examples. There are one or two European markets that we're looking at very carefully, I hope that gives you some perspective.
Your next question comes from Tahira Afzal with KeyBanc. Your line is open.
Hi, KR, team. Nice to see you back on track. Congratulations on the installs.
Thank you.
KR, first question for you. If you sit one year back from now and you see where your backlog is right now, where does it surprise you in terms of where the orders have come in from?
We are extremely happy with the diversity that we have. Both geographically as well as strong in various sectors. What we're excited about, without getting into the details right now, Tahira, is that we have newer opportunities that we did not have before, so these are new avenues that we can explore. Equally thrilling for us is the three sectors we have talked to you in the past, which is edge data centers, healthcare, and retail. They continue to be very strong for us, and we expect that to grow. I think you heard the comments from Matt, there is an opportunity here for us to make a big difference in the data centers, not just in the U.S., but because of the Power Tower, maybe in Japan, which could potentially be a very large opportunity.
When I look back at our backlog, very strong in the areas that we've been traditionally strong. Added to that is the geographic diversity, on top of that, you add some new areas we are super excited about growing in.
Your next question comes from the line of Michael Weinstein with Credit Suisse. Your line is open.
Hi, guys.
Hi.
Hey, when you look at the backlog that would have existed, or it did exist back in the fourth quarter of 2017, at the end of 2017, and that converting to 809 acceptances this year or last year, what can we infer about acceptances for 2019 versus the backlog that you now have of 1,400?
Yeah. Look, Michael, I think that's a great question. There is clearly some backlog that we have, that we exited 2018, that will end up being early 2020 revenue. As we look through, part of what was in the 809 and part of what we expect in 2019, there are additional incremental bookings that will come in in Q1 and even later in the year that will end up being revenue that we will achieve in 2019. A great example are our Korean orders, for which usually they come in one quarter and are shipped the next quarter and recognized as revenue as an example. We have a pretty good visibility on what's going on in Korea for 2019.
I think the walk away from that is, as we said in our prepared remarks, our backlog is 70% north of what we had in our acceptances in 2018. We're getting, as we pointed out, a little bit better at our execution on the U.S. C&I, and we have the international where we don't do the installs. It's not just Korea, also for Japan, and regions as well. We think it bodes well. We're not providing guidance for the full year, but as we did say in the remarks, by the second half of the year, we certainly think we'll be in that 25%-30% revenue year-over-year growth range. Even expecting some contraction year-over-year in the ASPs, we think that backlog bodes well for us to be able to achieve something approaching our longer-term 30% revenue growth rates.
I think in summary, what Randy's saying, based on our exiting backlog and what we expect to be booked, is the basis on which we expect our second half growth to be the numbers that we gave you. We feel very good about where we are sitting right now.
Your next question comes from the line of Colin Rusch with Oppenheimer. Your line is open.
Thanks so much. Guys, can you talk a little bit about how big a percentage of the backlog in the 2019 unit acceptances you're expecting to come from Korea?
I want to make sure I understand the question.
Yeah, I just want to understand how much of the backlog that you have is from the Korean customers and how much of the growth and the total acceptances you're expecting in 2019 are coming from overseas. Then I have a follow-up.
Look, the Korean backlog that we have today, I don't have that number in front. Matt?
I have what it is today.
Give an approximate
it's a high single-digit number.
High single digits that we have in backlog today.
As a percent.
Understand, based on what Randy said, the Korean backlog comes in and out with a pretty short shelf life.
Yeah.
Don't take that number as indicative of what the business will be.
I guess the clarifying question, which I think is related here, is how much of the acceptances you're expecting to have in 2019 with the unit growth are you expecting from Korea?
We think our Korean revenue in 2019 will be north of 25%, maybe approaching 30% of 2019's total revenue.
That's super helpful. I just want to make sure I understand the shipment terms really clearly. You recognize revenue when the units get to the port in California, then the payment terms are 60 days after they arrive in Korea. Is that correct? I just want to make sure that we understand the working capital needs accurately.
Slightly changed from that. As you know, we build our units in Delaware, we don't ship them across the country.
Okay.
They're getting close to Delaware. Yes, once the title transfers, once they're at the port and put on a ship. Yeah, go ahead, Matt.
Yeah, I was just going to jump in just because I live this every day. Our terms for acceptance last year were departure from U.S. port. In 2019, it's arrival at Korean port. Simple as that.
Yeah.
I apologize, we've run out of time. I will now turn the call back over to the presenters.
Yeah. We want to thank you for the time, and appreciate your support, Bloom, and we look forward to talking to you throughout the quarter on the next call. Thank you.
I would like to add to what Randy said and say that, look, our solution of always-on low carbon footprint is more important than ever, not just in the U.S., but everywhere else. I'm super excited that in more and more parts of the world, we are beginning to be mainstream in more and more sectors. We look forward to growing this company with you, and we are constantly making sure that while we are meeting our short-term goals, we are focused on building a great company and doing the right things to build a great company. Thank you.
This concludes today's conference call. You may now disconnect.