Good morning, everybody. Thank you for trekking out to this strange part of the world where normal businesses live. We have quite a lot to get through this morning, and here is the agenda. We've conveniently put page numbers, so if you want to jump back and forth, you're very welcome to do that. We're gonna go through a little bit of a sort of like: Okay, what's the environment in which we've delivered these results? What are the results? So that you can actually, hopefully, fully understand that. A little bit more about the investment case, because I think our business has continued to evolve, and I think it's important that you understand why anybody should invest in Next 15. And also, you know, some updates on what we're doing from an operational and strategy perspective, and then, not surprisingly, a bit on the outlook.
The first slide here talks about the market. I think before I jump into that, it's important that you know that this is possibly the most challenging environment in which we've run a company. I'm old, and I've done this for a very long time, and I've been through recessions, and I've now been through a pandemic, as you all have. We are dealing with the aftereffects of the pandemic. We are dealing with, you know, the end of free money, crazy high interest rates, challenges still with the supply chain. For many businesses, we're still dealing with the hangover from a tech perspective of COVID, where people bought way more technology than they were expecting to buy, and now they don't need as much. We're dealing with that, sort of like, it's almost like the snake digesting the deer.
You know, it'll eventually want to eat again, but for a while, it doesn't need to. And we've done all of that, and yet, actually, we've managed to continue to grow. So from my perspective, the fact that we've accomplished that and the fact that we've managed to maintain the margins, the fact that we haven't sort of fallen into some kind of pit, is, I think, huge testament to the way that we run the business, huge testament, really, to the businesses that we've got. I don't think we can take a lot of credit for that. I think they can take a lot of credit for that. It's underpinned by some fairly simple things. We do a really good job for clients, and they stay with us. If you look at our client retention, it is phenomenal.
We've talked about the fact that the technology industry is challenging and has been challenging. I actually think we're going to see a small improvement in the second half, and I think next year could actually be another year of quite good growth in that area. But if you look at our technology customer base, they have stayed loyal, even though there have clearly been areas where they've wanted to cut spend. And in fact, if you look at our top 20 customers, many of whom we've had for a ridiculously long time, overall, their revenues have grown in the period. You know, you've got a few odd exceptions, people like Google, whose revenues have gone backwards. I think Google was quite public about saying: We're gonna just save everything we can save anywhere and invest in AI. And unfortunately, that applied to us as well.
We did see some areas where their spend increased, but overall it went down. So that sort of cornerstone of our organization is very firmly placed, and I think is something that I'm not sure the shareholders always fully understand. I think they sort of look at the complexity of our business, and they sort of lose that data point in the process. I think, you know, there are other things that have been going on in the market which are more subtle. The demise of the cookie has accelerated a bunch of things, but really all it's been doing is accelerating a digital shift anyway. So almost every purchase now, in some way, shape, or form, has a digital component to it. People look at something on Instagram.
Certainly, if it's one of my kids, that's the way that the purchase process starts, and then they eventually buy it. They may buy it online or they may buy it in store, but there is a digital influence to that purchase, and there is a digital customer experience to that purchase. There just always is, because there's some follow-up from the retailer or whatever it might be, and that applies now, not just in the consumer world, but very much in the B2B world. And so that whole shift is changing where money is spent.
It's moving it from a world where people would spend a lot of money on brand building, you know, sort of the long tail, if you like, of marketing, and saying, "Right, we're gonna concentrate a ridiculous amount of money into that short window where people are looking to buy something and they purchase it." And that is actually kind of a significant restructuring of the industry that's taking place. And again, I'm not sure that shareholders have seen quite what that means, because I think people see structural shifts like, okay, TV advertising is, you know, on this significant decline and the rest of it, but it's why those things are happening. It's not just because people aren't watching TV in the way that they used to watch it.
It's because actually the purchase decisions are occurring in a very different way, and people are realizing that that last minute on Amazon, when they were gonna buy this, but then, oh, I just decided to buy this because the recommendation engine pushed me to something else. They're realizing that that's not just an Amazon thing, that's a online thing, and if they fail, you know, this is where businesses are making or breaking their numbers, is because somebody clicked this versus clicking that. And so that whole issue of where, you know, the decision point is now becoming absolutely crucial, and they're realizing that whereas before that decision could have been swayed by just doing a really good build-up in brand building, that loyalty just is not there anymore.
Other things to note, really are just the environment in which businesses like ours are actually operating, and that last year, wage pressure was a thing. You know, quiet quitting was a thing. There were a whole load of things on the whole staffing side of our organization that we were trying to navigate. A lot of those pressures have pretty much vanished, to be perfectly honest. It seems that as people have returned more to in-person, if you sort of wander through the building after this, which you're free to do, you might get kicked out by some people. Certainly one part of the building I'm not even allowed into. You'll see a lot of people in the office, and I think that has had something to do with it as well as people have sort of realized, "Okay, I actually am attached to a company. My loyalty is now starting to reappear within those companies." And so people's sort of fluidity of job movement, which seemed to be certainly there during COVID, has very much dissipated.
You know, I think a lot of the reason that we've succeeded is because we are a collection of specialists. I think we get a lot of feedback out of investors, which is: You're complicated, you're difficult to understand. You have all these businesses. Yes, we are, and that is why we're good at what we do. It's because we actually have businesses that are incredibly close to the problem that a customer has to solve. They can adapt very quickly. We don't have to go running around saying to them, "You need to cut this cost or invest in this product area," or whatever.
They do what is right for their customer, and I would much rather we have the problem of trying to explain our business to you than we have the problem of having some monolithic business that our customers don't understand. You know, they understand our business because they understand the problem that they're trying to solve, and we have a very specialized part of our business that will go and solve that for them. And then when they don't need that, we can get out of the way. And that, to them, is a perfect supplier. And increasingly, with technology, that model is very effective for them. It used to be that big organizations in particular, liked the laziness of saying, "I'll have one company. If you can just do everything for me, that would be very helpful." That was the proposition that the likes of WPP and even the large consulting firms tried to pitch.
With technology, that's becoming less and less of a glue because the laziness doesn't matter anymore. It's much easier to manage a portfolio of specialist businesses than it was. So, therefore, if we're doing a really good job for somebody somewhere, and WPP is trying to make a bid to procurement to say, "Well, you know, we can do this package thing that gives you this much benefit," that part of the business will fight for us and say, "No, I don't need to change. I'm very happy with them. They're doing a really good job for me. Saving 2% just isn't worth the hassle, the change, the disruption to the business and all the rest of it."
We are a very diversified group. We'll get on to that a little bit as we go into some of the issues to do with megatrends, that we're sort of building the business around and the sectors that we're building around. The agility of the business model. As I said, people are just very close to the problems they're trying to solve, and that's super important. You know, I think, again, that's part of the sort of investor case that gets lost, is that people don't realize that actually being an agile business is so important these days. If we cannot literally change the product, change the service quickly, then, you know, customers will move.
You know, it's a bit like I was saying about people changing decisions on Amazon. You know, it's like if you cannot offer the right thing, then you're creating a point of weakness, and so our agility is hugely important to me. Lastly, you know, this focus on organic growth, you know, and the disciplined approach to M&A. You know, I think we got criticized for the Saatchi bid, which, you know, I think we've talked about being relatively opportunistic and really being focused on a few of their specialized businesses that we really liked. I don't think you will see us going off and doing anything like a Saatchi bid.
I don't want to go through that nightmare again, and I know that we can do a lot of really interesting things without having to, you know, waste the time and the money that went behind that kind of thing. There are some other things that have been going on in the business behind the scenes that you may not have noticed. You know, we continue to do small bolt-ons. You know, we've added some into our Palladium business, we've added into our Brandwidth business, and these are really just bolstering existing capabilities. You know, we're typically paying 5-6x EBIT for those businesses, so they're very easily accretive, don't use a lot of cash. You know, they just continue to help us build real, good, quality, specialized organizations. In the center, Peter's team has done a phenomenal job of just really driving the efficiencies.
A lot of the efficiencies that we expected to come out of the Engine acquisition, you know, moving into this building, getting all of our real estate really much more concentrated in London, for example, has yielded a big saving that will be a lasting saving for the organization. But it's also, we had duplicate head office structures between the Engine business and ours, and, you know, there is a little bit more we can go on, both of the real estate front and on some of the operational savings, but I'll get into how I want to spend some of that money a little bit later. On the other side of here, on the slide, is the client wins. We do continue to win some really good clients. These clients add up to sort of roughly 5% of our revenue, you know, on an annualized basis.
So we do continue to add some very good clients to the mix. And if you look at that, you know, it's a pretty broad spread. Yes, you've got a pretty well-known technology brand in there, but you've got, you know, Morrisons, the Department for Education. You know, we had nothing in public sector until we did the Engine transaction. Now, that is becoming quite a meaningful part to our business, as you'll see later. Financial highlights. By this point, you've already read all of the numbers, so I'm not gonna go through them in detail. I think, you know, we had this distortion that we did flag this time last year. You've probably forgotten about it, which was we got this sort of additional revenue coming out of the large Mach49 contract win. If you were to strip out that GBP 5 million t hat sort of added to both revenue and profit, quite frankly. It literally was a complete flow through.
EPS was always going to be down this year to some extent, because of the tax changes in the U.K. I'm particularly proud of the fact that we're managing our balance sheet well. We've gone from. Remember, during COVID, people paid in two seconds, and people have definitely returned to a more normal paying process. But our guys have done a, I think, a pretty good job in managing that shift. The interim dividend is only up a modest 5%. We're never going to be a yield stock. I think we would have put it up more had we not announced the share buyback that we did today, and I think that's just down to competing uses of capital.
At that point, I'm going to hand over to Peter, and he can give you the wonderful color on the numbers.
Indeed, indeed. Morning, everybody. Yeah, Tim's covered off some of these, but, yeah, revenue was up 5%, 1% decline in organic, which would have been sort of up a little bit if it wasn't for the Mach49 situation. Again, operating profit down a little bit, distorted by the Mach49. The only real sort of highlights or lowlights maybe is tax is now 27%, mostly because of the government going from 19%-25% in the U.K., but also a high proportion of our profits coming out of the U.S., which typically has slightly higher tax rates. We've increased the dividend a little bit. We'll talk a bit more about sort of capital allocation later. Revenue bridge, yes, at 1% decline, we had 4% growth from acquisitions.
That was mostly Engine, which we acquired in March last year, so we had six months in the sort of half year this year versus five last time. So that's about sort of GBP 7-8 million of that. And then, as Tim said, we've done a few bolt-ons, which have sort of, you know, continued to grow, and a little sort of modest tailwind from acquisitions, so, sorry, from Forex. On segments, so we had pretty good growth from inside. Again, that's where bolt-ons tend to be concentrated. Savanta, we've grown from GBP 1 million revenue when we acquired it eight years ago to roughly around rates of GBP 50 million. Good organic growth, but lots of bolt-ons. I think we'll continue to grow.
Canada is a market they would really want to grow into, so hopefully, you know, a couple of bolt-ons in that area in the next sort of month or so. On margin, down a little bit, that was really from investment in Plinc, which is our sort of data science business. They've developed a couple of products, and we've hired someone to really drive a sales function. So we think there's a lot of opportunity with, you know, particularly retailers, to get some sort of products into that space. So I think that investment will continue for the next six months and really start paying off next year. So I think the margin will be slightly below 20% this year because of that investment, but should get back up above 20% next year. Engine, quite a broad church in here. MHP has done very well.
Happy to say that because they're in the room, but they've had a good, strong period, M Booth has done well. Some of our sort of Google revenues, Tim highlighted, is off a little bit, you know, as they've gone from, you know, sort of brand building type work we typically do for them to AI. But overall, I was very pleased that we retained margin at 20%. Again, I think that margin should be a little bit better in the second half. We tend to have a stronger second half of the group, with September, October, November being our strongest months overall. Customer delivery, we've got two sort of standout businesses in this space. Activate continued to do very well. I mean, they had a phenomenal period during COVID, where online lead generation, you know, grew like absolute weeds.
We were slightly sort of nervous, would it be, you know, quite a sort of a down period for them? But actually, they've maintained their revenue up against, you know, people like TechTarget, who've had a very strong organic declines in Media Logic. So, you know, that business feels a bit more optimistic. In fact, we had a sort of very optimistic email from them a couple of days ago. So again, I think we're beginning to see, you know, signs of growth back in that business. And SMG, who've had a phenomenal period since we acquired them two, three years ago. They, you know, won Morrisons, which is an interesting from an accounting perspective, because the way that contract works is they get a monthly retainer, which pretty much pays their costs.
And then in the first year, we're going to account for all their sort of commission revenue in a lump sum, because, you know, when we started the contract and started, you know, the financial year, we didn't quite know how they were going to do. As it turns out, actually, the business has done better than expected. We had a very positive board meeting with them yesterday, and, you know, I think there'll be a nice, maybe, sort of couple of million sort of benefits in the second half just from that contract win. So I think, again, customer delivery, I can imagine, will be second half weighted and hopefully the margin up a little bit in the second half. Business transformation, a sort of tale of two halves, really.
We have two businesses in that space, Blueshirt Group and Palladium, who have both suffered from a bit of market weakness. Blueshirt Group is very much focused on tech IPOs. There weren't a huge number of them last year. We expect, you know, on the back of Arm and a couple of other sort of U.S. tech IPOs, which have seemed to have gone well, that market will open up, but, you know, that won't really come into being till next year. So again, I think that part of the business will be quietish. Similarly, Palladium, which does sort of DD for PE firms, and again, that's been a bit quiet.
On the other side, Transform, which again, you know, joined as part of Engine, had a phenomenal first half, very much on the back of the Department for Education win, and, you know, we're beginning to do good growth there. And also, they've sorted their margin out. When we acquired it, it was roughly a 5% margin business. It's now sort of 15%. Really did some sort of fairly basic sort of housekeeping and getting sort of better systems in. So that part of the business is doing very well. And Mach49, obviously on the back of the contract win, has done exceptionally well, very high margin and, you know, I can see that continuing to grow and deliver well in the next couple of years.
Finally, head office costs, I think I mentioned we've done a good job in terms of reducing them as a percentage of revenue. Cash flow, we've slightly sort of tweaked this because acquisitions in statutory accounting for acquisitions with earn-out payments have to go against the normal cash flow. So we think this is a better comparison in terms of how we've done. So we had a small working capital outflow, which is very much based around bonuses. We accrue bonuses at the end of January and pay them in sort of April, May, and then it start accruing as the year goes on. So that's probably the biggest impact here. We typically have a little bit of working capital outflow, normal trading in H1, and, you know, we're working very hard, as Tim says, to kind of pull that back in the second half.
So it will be less by the second half, genuine wages and tax, I said, has gone up a little bit. CapEx, I think, is fairly modest, and the acquisitions, because of the share price, paying cash for earn-outs rather than shares, you know, that number is a little bit higher than we alluded to six months ago. And I think overall, with, you know, small bolt-ons, et cetera, I'm expecting numbers to be about GBP 70 million for the full year. But even with the buyback, which I think will probably be about GBP 10 million, this side of the financial year, we should be sort of positive net cash for the year end.
Yeah, as we understand it, the buyback piece, we'd like to do as much as we can, but it, there are limits on how much you can buy, given volumes and so on. So realistically, only about GBP 10 million will come out this year.
This is our sort of estimate of future earn-outs payments. It's come down by $20 million, which is sort of two factors. One, we made our first big payment on Mach49, which was about $45 million. So that's sort of the impact on the reduction in FY 2024. But also, we're now assuming that all future payments are in cash, which has increased the number by $29 million. So those two impacts sort of offset each other.
Obviously, if our share price improves, we can decide to go back to paying the component in shares.
Finally, you know, there was a lot of debate around the board in terms of our capital allocation policy. I think, you know, we're gonna continue to prioritize internal investments and the obvious area is AI, which I'm sure Tim's gonna talk about as we go on. You know, the policy here is we want to invest both to, you know, realize efficiency savings, but also to create new products. You know, we're looking to have a sort of capital markets event towards the end of October, where we're gonna actually demonstrate some of the, you know, products and efficiency savings we're realizing from the businesses already. You know, it's still obviously early days. You know, we're still pretty allergic to debt. I don't think there's any banks in the room, but I think they will be pleased to hear that.
So whatever we do in terms of capital allocation, you know, 1x EBIT is already somewhere Tim and I start switching. In terms of bolt-ons, typically, yes, we, we can pay somewhere between sort of 5-6x EBIT. Also, our model is that we typically pay, you know, 50%-60% up front. Yeah, we like to use shares if we can, because it, you know, gives a little bit of loyalty back to Next 15, but obviously we'll have a sort of view on that. But, you know, I think we're pretty disciplined and focused around a few key businesses like Savanta, like BrandVue, good operational teams who've actually used to sort of taking these businesses on. We're still keen to look at strategic acquisitions.
I agree, probably not at the M&C scale, maybe slightly more at SMG scale, which was a, you know, GBP 20 million revenue business, but, you know, was clearly in the sort of retail media space, which is, you know, sort of high growth. I think we recognize it's sort of harder to do in the current markets, but not to say we won't be scouring the markets for something which, you know, is slightly more needle moving. In the absence of that, and obviously particularly with our sort of current share price, you know, we are looking at share buybacks, and our head is at, you know, doing GBP 30 million over the next 12 months. As Tim said, there are constraints. We can only buy up to 25% of our historic liquidity, which at the moment is fairly modest because of, you know, the summer period.
So I think we worked out it was something like sort of up to 30,000 shares a day. So GBP 30 million is gonna take quite a while to get through there, unless the liquidity doesn't improve, and I think we're gonna be flexible. I think in my head, we're gonna do GBP 10 million this side of the financial year. Maybe that moment to sort of reassess, depending on where the share price is and where our priorities are. And similarly, we're gonna be looking at these earn-outs and whether, you know, again, it's the right thing to do to pay in cash as we did this year or to use shares.
Yeah, I mean, these are very much competing uses of capital, and we're sort of balancing one against the other. And, you know, I think the share buybacks is if we can't find a good use for the cash in other ways, then yes, we will do this. If, you know, when we did look at a couple of acquisitions earlier this year, and the multiples were sort of like 12x and 13x, they were great businesses, you know, et cetera. It would have been quite, you know, some good things for us to own. But then you look at it versus saying, "Well, I could spend that money and buy a portion of a company I already know. In fact, I know it really well. You know, I know how it's likely to do. I know I don't have to pay an earn-out on it," you know, blah, blah, blah. I mean, you sort of arrive at the conclusion that it's very difficult to justify with the share price where it is, buying something at 12x, unless you think, yes, okay, this is a meteoric growth and, you know, transformative from a margin perspective. One of them wasn't. One of them was only a 15%-16% margin. High growth, but you know, relatively modest margin.
And that's why the things like the share buyback just makes sense at this point. But, you know, they may well time out. Peter and I, and JP, in fact, all read Outsiders over Christmas last year. Well, that's how exciting we are, and I don't know if any of you read it, but it's a great book, and it's all about capital allocation.
Again, it's how exciting we are that we read that over Christmas. The very simple message that comes in there is: Use your share price when it's strong to go and invest and buy into things, and when it's weak, buy your own shares. I mean, it comes across loud and clear. They essentially follow nine different businesses and all of whom were transformative, and within their capital allocation policy, that was an absolute rule that all of them applied. Warren Buffett is one of the cases that they follow in all of that, and, you know, they just couldn't be clearer on it. Just stepping into the investment case, you know, we think of ourselves as, you know, we do this sort of classic thing that, you know, all businesses do, which is, you know, the what business are we in?
And at a very simple level, we're in the business of maximizing potential for our people and our customers. You know, it's we are helping them grow their sales, we're helping them grow their market share, their share price, you name it. That's what we're in the business of trying to do. And for our people, it's a case of trying to maximize their potential, you know, from a skill-based point of view, from a career point of view, from hopefully from a financial point of view. And what that really translates into is what we believe is a growth consultancy. And I think hopefully that better describes that whole concept of what a growth consultancy is, because I think some people have kind of said: "Well, I don't really know what that truly means." From our perspective, it does break down into four business areas.
None of this is particularly news to you. You know, there's the insight business, which is increasingly more about how we use data to solve problems for customers. Insight, in fact, is possibly starting to become less and less the right name for that, because increasingly, customers are actually coming to us and just saying, "I need the data. Can you give me the data? Yes, we may need you to do some insight work on top of that, but actually, the data is the thing that we really need, and you are really good at finding that." And our market research chops in that area, especially as we shift into AI uses of data, are becoming of increasing value. If you wanna find out who's gonna win the next election, you don't go and interview 56 million people.
You find a subset that is believable, and the same is true for any form of data that you want to apply something. If you want to train a model on something, you don't need the entire universe of data, you need the right subset. In customer engagement, you know, it's about optimizing brand assets, but it's really about customer experience. And that's really where that whole area of our business is being pushed. In delivery, you know, this is, there are two sides to that business for us, really. It's demand gen, and it's connected commerce. But again, it's all underpinned by data. I mean, that business area, there is literally nothing that a customer is buying from us that isn't deeply embedded in data. And then on business transformation, this is where we're typically competing with the very large consulting firms.
You know, we are brought in to solve a value problem. And in fact, if I sort of click on, the next slide is looking at business transformation. And really, you can think of it as these are all businesses that, in some way, shape, or form, are tasked with value creation. You know, it's a very fundamental level. This is not sort of like, "Can you help us move our share price a little bit?" This is, "No, we need you to fundamentally position us in the right way so that we are maximizing the value of the company that we are building," or, "You need to help us build the right company." You know, and we are literally competing with the likes of McKinsey and Bain on designing and building new companies and new products, and that's on the Mach49 and the Transform side.
On the Palladium and Blueshirt side, that is more, "Okay, let us find the assets of real value within your organization and make sure those are the things that are really seen and understood, so that we can truly maximize the value from an external perspective of your company." In terms of mega trends, this, I'll confess, is something we entirely stole from private equity. We realized that we are a sort of a private equity firm in disguise. We had this fascinating conversation with Bridgepoint about this topic, and we realized, actually, we do exactly what they do, which is we find things that we think are the things that are gonna be really good long-term drivers of business growth and build businesses in those spaces, and these are them. You know, data-driven decision-making, that's our insight business or our data business.
Customer experience, as I mentioned, is really our customer engagement business. Sales engagement is our delivery business, and digital transformation is our business transformation area. Those are all areas that we know are going to continue to just naturally grow. Customers are going to buy more of all of those things over time. Our challenge is designing and building and executing well in each of those areas, and if we do, there is no reason why we're not successful. The other axis for that is: Where do we do business? You know, we've historically been known as a business that is really rooted in technology, and technology very much remains the heart of our company.
What's pleasing about this breakdown is it shows, yes, technology was a tough place for us in the first half of this year, but we made up for it by, you know, growth in financial services and public sector. That sort of, again, kind of speaks to the model that we operate, and if we start to see technology rebound, as I suspect it will certainly next year, then that should put us in a very good place. What you're not seeing is sort of aerospace, defense, you know, other areas, transportation and so on. You know, we want to stay very focused. You know, part of this whole specialist thing is staying in certain industry sectors and certainly very specialized product sectors. If we start to spread ourselves too thinly, then we start to lose value to the customer.
One of the things they love about us is there is zero learning curve. We walk into a customer, and we understand the problem that they're trying to solve very quickly. We can add value very, very quickly, and unfortunately, that, these days, if you can't do that, you are not going to succeed. The only way you are gonna succeed is by offering it at a ridiculously low price, which we don't want to do. Moving to AI, we are in a position where we could, and we could have actually in the first half of this year, we could have delivered better profits. We could actually push our margin up over the next few years by two or three percentage points quite easily. I am not going to do that. I want us to invest more into AI.
Now, that's not to say that that is all that we're going to invest. We are basically redirecting almost any money we can find out of almost any pot into this. You know, we've been working in AI, you know, both from a customer perspective and from an investing in products perspective for quite a number of years now. Generative AI has essentially just accelerated a whole load of that across the business. It used to be in just very specific areas. Now, it's literally everywhere within the organization. The spend is split into really two different areas. One is just efficiency, accelerating existing processes, making it easier for us to provide a product or service to a client so that they can do it more quickly, maybe on a more self-service basis.
And then there is, let's create completely new things, create new products, new services, that create a completely different solution for customers than is currently available. To give you a few simple ideas of stuff that's been done, we have, like, a lot of services, businesses, challenges around, you know, how do we make sure that all of the content that we're producing for brands is on message, is following the brand guidelines and all the other stuff? You know, and if you're producing literally thousands of pieces of content a month, that's quite hard to do, especially when they're being created by relatively young people who are notoriously lazy and don't do everything that they should.
What we've created is essentially a version of ChatGPT that, you know, it's a private instance around each individual brand for a customer, and you can literally test every piece of content against all of those guidelines and make sure that it meets the guidelines and, you know, it will tweak the language for you and put it into the right place. That's taken a problem and actually turned it into a benefit. Now, the customers are saying, "Actually, can you run this additional content through that for us? Because actually other content that we're not producing, can you make sure it's on message and all the rest of it for us?" It's a small but nice incremental piece of revenue that we're getting out of that.
We had a product called Prism, which was built by the Palladium guys, which was to create a self-service due diligence tool for private equity. So the private equity would, you know, essentially go and look at more companies, and that would give us amazing visibility on the types of stuff that they were looking at. And then it also became something where they were like, "Oh, we should look at five companies, not two," because it was easier for them to do. Now, we've embedded AI into all of that. So what it does, it does a whole bunch of different things, but the simplest thing that I can explain is now it's become a live tool, so it becomes a tracker. Effectively, somebody says, "I was interested in this company. I can now see how that data is changing."
So that thing that looked quite an interesting investment three months ago doesn't look quite so interesting now. That thing that didn't look so interesting three months ago looks actually quite interesting because it's constantly refreshing and finding new due diligence on behalf of private equity firm. And so, again, that's just a nice, easy thing to do. With Agent3, you know, they do an awful lot of email marketing as a part of ABM, and a lot of it is because it's B2B tech stuff. It's pretty heavy data, you know, so you're going to be pretty brave to click that email and open it. And what they've done is essentially created a way by... What it does, it takes all of that data and puts it into a video.
It automates the creation of a video, and their click-through rates on that have skyrocketed because it just creates a very short video that explains whatever it is, this complex topic, piece of technology, whatever it might be, and people are much more inclined, it seems, to do that. I wouldn't be, but then I'm not a B2B tech buyer. But, you know, those are just some small, simple examples of the types of stuff that's going on. I want to give you a sense of the ambition. In the last five years, we've more than doubled the company from a revenue perspective. I believe that we can do the same again.
The reason I think we can do the same again is because we're actually in four areas that I think are going to continue to grow, partly because of the mega trends that I talked about earlier, partly because of the customer focuses that I talked about earlier, that give us a massive opportunity. Data is going to be a massive growth business. If I'm really honest, if it's not for us, it will be a huge missed opportunity. It is a completely new market in some ways, because of AI. The transformation that that sector is going to go through is quite interesting. The market research businesses in that space, if they do not rethink of themselves as data businesses, then they are going to struggle, quite frankly, because AI is going to automate an awful lot of the market research processes.
It's going to drive a lot of value out of that, but the underlying data is going to get, ironically, more valuable. We're going to see a decline in value on one side and a massive increase in value on another. On customer engagement, I think there's more modest growth opportunity there, but it's getting more and more and more digital to this point about customer experience and the digital sort of interface that goes on there. So I still think there's some good growth there. Customer delivery. IBM used to take, and still does actually, a ridiculous number of customers to the Masters golf tournament in Augusta. Costs them an insane amount of money.
We proved during COVID, because we had the data to prove it, we generated more value out of the $8 million they spent with us that year, and generated more sales and all the rest of it, than they generated by the more than 10x just for that one event. Companies are realizing that, yes, that stuff has a role to play, but it is so much better to spend the money on proper data-driven customer engagement work. Interestingly, you know, TechTarget had a really tough time in this last six months there. I think they had an organic decline of 25%, which I was kind of blown away by.
In talking to both Agent3 and Activate, which are two other businesses we have in that area, both of them came back and said that's because their product, they've moved so aggressively to SaaS, and the customers want TPaaS, which is, you know, technology and people. So they want that combination because it's-- they find actually the TechTarget product too difficult to use. It's producing sales leads, and they're not completely convinced those are the right sales leads. They're not completely convinced those sales leads are ready to be given to the sales team and the rest of it. They want a human layer on top of it. Now, I suspect in the next four or five years, that human layer will get eroded, but in the short term, it's still the way to deliver it. But it is a massive, massive growth area.
If I'm completely honest, I could have turned that number into GBP 500 million. It's the market opportunity there is just huge. Business transformation, similarly so. I mean, you know, if we're competing with McKinsey and Bain and the rest of it, you know, it's really just a function of us doing a good job, executing well, and if we do, we can easily deliver that kind of growth. In other words, you know, I'm not sitting here going: Yes, okay, market conditions aren't perhaps as rosy as they were last year, but long term, the potential for us is still huge. Quick update on strategy. We will remain specialized. You're not gonna see us change the structure of the organization. There'll be some small areas of simplification, but nothing major. AI is going to be everything.
You know, if we can find a penny on the floor, it'll get spent on it. You know, we do keep pushing up within organizations. It's the thing that we could do better, but, you know, with relationships that are starting to develop through places like Mach49, that's getting easier and easier. M&A will still be focused on the mega trends I discussed earlier. AI data sets are going to be of massive importance, and I separated the two things out because actually, the investment in AI products is one thing, but if we don't have the underlying data, we definitely will not achieve what we need to achieve. And then, obviously, we've kind of introduced this new element around capital allocation. In terms of outlook, you know, we do continue to trade broadly.
I think that was the word that everybody settled on, in line with management expectations, despite all of the headwinds. You know, I think we will still grow this year. Not perhaps what I would have liked us to do, but, you know, I think given the conditions, I think it's... Especially with what's going on in tech, I think that's still a pretty good outcome. You know, we do still see good progress. You know, one of the things about having a portfolio is you can really see where the money is. You can see which areas of business are doing well, where customers are directing money, and where they're pulling money. You know, it gets masked. As soon as you start to push things together, it's harder to see what's going on.
We can really see it, and it's clear that there are still some really exciting areas of growth, and we just have some very good businesses. I mean, SMG in and of itself is a phenomenal company, an absolute phenomenal company. It is going to continue to grow at incredibly good rates and deliver very good margins, and that's just because they're a really good company in a really good space. If anything, the market is just coming to them. We do add new clients. I talked about that. We continue to run the balance sheet cleanly. You know, as Peter said, you know, he and I get very twitchy if we get even close to 1x EBITDA, and I don't think we'll ever get that close, and we're just gonna spend every penny on AI. That's it.
If I haven't said it enough times, that's what we have to do. We can still do good, strong margins. You know, I'm not saying that we're gonna take a 20% margin and turn it into a 16% margin. I think we can stay at and maybe even still actually increase the margins over time. But I'd rather say, let's continue to deliver a good, strong margin, and if we have excess capital or excess profit, effectively, then let's reinvest that into the organization in the right ways for the long-term growth of the business, rather than sort of artificially raising the profit margin, which in the short term would be good, but in the long term would be, I think, a bad move. We're done. Questions?
Crystal, some of the things you were saying about the AI, absolutely fascinating. How siloed is the development? I'm thinking, you talked about taking complex information, pulling videos. Can that be used throughout the rest of the business?
So, so JP, who's sitting over there, Jonathan Peachey, he can stick his hand up. He's actually over... So he's driving all of our AI initiatives across the group, or is sort of making sure that everything gets coordinated, and is doing a lot of really good sort of. We're sort of trying to do two things. We're sort of really pushing the companies themselves to do their own level of innovation. And then, at the center, what we're trying to do is challenge them with ideas. You know, so for example, we had a summer of code this last year, where we hired a whole bunch of ridiculously smart but ridiculously young, and almost none of them were from England, people, which is really sad. I think there was a one person from England, JP? Yeah.
Eight, nine people from India and one person from China or something. It was kind of like, okay, where are the skills then? Not in England. And we're extending that process, but that was a way for us in the center to just kind of, like, take ideas, throw them in, prototype them and see what could be done, and then throw them into the businesses and say, "Right, you go play with this, see if there's something that can come out of it." And we'll just carry on doing that. But what we are doing, and we're getting really great reception on, is, "Okay, you're all—you guys are all building something that's kind of similar.
Let's put all of the code into GitHub so that you can all see what's been built, and then you can actually, rather than you having to build from scratch, you can basically take whatever it is that you need and create your version of it." You know, like that product I mentioned about the, you know, brand guidelines and brand consistency and so on. That's a perfect example of that. It's like, there's no reason why that has to live within one of our businesses. It's not actually that hard to build in terms of, like, the basic code. What you do need to do is do all the work to ingest all of the brand content that's specific to that particular client.
Within the constraints of something like ChatGPT, where you've got a limit of 25,000, you know, it's, it sort of does create some complexities because you then have to try and figure out, like, which content do you ingest and some of those kinds of things. So that's where actually creating the code part of it is only one part of the challenge. So let's just, like, make that really simple. Let's just, you know, do it once and then use across the organization.
I do my other two more-
Sure.
The tech trends, you were saying you could see things getting better in the second half and into next year. Is that just lapping the poor performance, or are you actually starting to see some things? And then my third question was just about new business and the situation, what the big picture is?
Some of it is lapping poor, poor performance, if I'm being very honest with you. Some of it is also AI, which is starting to really impact in a good way. You know, semi had a really good time during COVID, you know, massive supply chain shortages and blah, blah, blah. So they just had an abnormally great market. And that has sort of hit. In some areas, has completely hit a wall, but in other areas, obviously, through things like NVIDIA, you know, it's just, like, pretty crazy time. But that is actually—I think people see something like NVIDIA, but then there's a whole layer of other small semi companies that are also starting to really pick up and do well out of it because they're creating other specialized chipsets, other GPUs, whatever.
So you know, I think there is a portion—you know, there's a portion of this is AI, there's also just a portion which is that this ingestion problem is starting to run its course, and companies are starting to reinvest in, you know, "Okay, I now have to update whatever it is that I was gonna update," or, "We have grown," or, "We bought a company, and we now need to put everybody on a single platform," just like we do. It's kind of annoying. We have to buy more economy, which everybody hates, but we do, right? You know, so it—there are just natural cycles that start to actually play into it, and that's really one of the reasons.
I think also... Assuming the trends follow what we're seeing and hearing out of the venture capital world in California, is that it was a ridiculously quiet period in the last, probably last 12 months, really. And they are starting to wake up, and they've definitely woken up on AI. They're, like, all over it. But they're starting to wake up on lots of other areas as well. And they are sort of a bit like the first-time buyer in the housing market. You know, if they're spending and doing, the whole thing starts to ripple through. So assuming that that really does come back in the way that it seems like it's going to next year, that could be very, very good for us. Not necessarily because we're, we are massively exposed to the VC spending, but more because of the ripple effect.
I'm sorry, what was your other-
It's on new business.
On new business. Actually, it's very healthy. I think the only thing on that is the time frames have definitely got drawn out. You know, and clients' start dates for things is where, you know... I mean, like, we have a number of businesses with quite significant backlogs of revenue, where the clients have said, "Yes, we're gonna do this, but we're probably gonna start this in Q4, or we may start this in the new fiscal year," type of thing. And so we just can't budget for that at the moment. It's sort of like we're sort of, we know we've won it, and we have, you know, the contract and all the rest of it. We just don't really necessarily know the start date for it. But overall, it's actually quite healthy.
You know, in some areas of the business, the competition is more, you know, you went from being two or three vendors were pitching to six or seven. But our win rate is still very good. So I, I'm not, not concerned about that as a, as a driver.
Ciarán Donnelly from Berenberg. A few from myself. I guess just kicking off, in terms of the shift you've seen in terms of short-term engagement solutions, conversion tools, one, can you just talk about whether you see that as kind of a perpetual trend, or is it transient, and is it linked to, I guess, business cycle? I.e., as we come down, people focus on the short-term tools, and as we turn, they might go back to four engagement tools.
That's a really good question. I unfortunately think it's a permanent shift. I think unfortunately, it's almost a bit like the political world is polarizing. You know, it's sort of you getting... You will inevitably have some brands in like a Procter & Gamble, Unilever, Coca-Cola, American Express. So pretty much all of those are customers. And they are. It's in their DNA. They just cannot think differently from that perspective, to a degree. But even those organizations have kind of also got a side to their organization that's gone really short term, that's like, "Almost what can we do this week? What can we do to sell more this week?
How can we shift the needle?" I mean, it's interesting because we see it through our SMG business, and we actually have all of their data through our SMG business, so we actually see all of the customer data that P&G and Unilever and all these guys have. So we can see also what promotions they're running and all the rest of it, and you can see the uptick where they are like, "Okay, we need to just drive more sales this week in this part of," you know, whatever country it is or whatever. And it's pretty fascinating. That, you know, that money is coming from somewhere. It has to come. You know, it's not an infinite pot of money, and I'm sure what is happening is they're just pulling from stuff that is less measurable and the rest of it.
But I don't, I don't think you're ever gonna go completely away from brand, but I, I think the, you know, short-term spend for brands, you know, if you go back 10, 15 years ago, it was a tiny portion of, of how they spent their money, for most. There were always some that were different, but for most people. Whereas that has now, you know, so much of the money has shifted in that direction, it's just never gonna go back, because it's typically very measurable. You can kind of look at the ROI on it, and you can kind of go, okay, you know, and you can literally see them move based on the data. It's like, okay, spend 10% more, spend 5% less, then 20% more, whatever. So they shift money in and out.
That makes it harder for us to forecast, but interestingly, thankfully, because we're starting to see the underlying data, we can kinda see why they're making those decisions. And that helps us navigate from that perspective.
And then just on the goal of about doubling the size of the business in five years, one clarification, I assume that's an organic goal, and there's no kind of inorganic baked in?
There'll be some inorganic baked into that. It's not going to be a huge component of it, you know, because we'll do bolt-ons. But it's... I can't bake into that, you know, a major acquisition otherwise I'm cheating. You know, so I'm like, "Oh, yeah, we doubled it. Okay, we did it." You know, it's like, it's like eBay used to say about winning. It's like, it's not winning, you just paid more than anybody. It's like, that's not really winning. No, I, you know, there'll be... It's incremental, acquisitions, not transformative.
Then, I guess, just in terms of the organic element, giving us a sense of how we should think about underlying market growth contributing to it, and then, I guess, in terms of AI, what's the benefit from efficiency gains versus new products? How should we think about in terms of those two?
You know, doubling it in that period, the efficiency gains sort of drive more the margin necessarily than they do the revenues. You know, I mean, there'll be some small incremental revenue that you'll get from driving a more efficient product. You know, it's like Prism. I can see customers are a little bit more addicted to the product than they were, and they're more inclined to use it because it's just... I can look at it every day, and it produces a different answer, right? So you sort of create more usage out of some of those products, but it's not transformative in the way that, you know, it would be if it was a radically new product. And I think that's the part that's hard to predict at this stage, and it's, you know, it's sort of like...
I think if we are able to produce radically new products, then I think we can do even better growth than this, I think is probably where my head would be. I think what I'm trying to sort of forecast with this is if we just execute well from an AI perspective, not brilliantly, but well, then we should be able to do this. You know, AI will have some negative impacts on our business, there is no doubt. It'll have negative impacts on everybody's business. It has to. I mean, otherwise, why are you doing it? And it's, that's the challenge for us is finding more positives than negatives. I mean, that I know it's obvious, but, you know, it creates lots of new ways to do things.
You know, one of the things that we were talking about is, we are this collection of specialist businesses, but that does make us, in some cases, a bit like an electrician, a plumber, a carpenter, you know, a roofer, whatever. And if you've got an electrical problem, you want the electrician and, you know, and so on. But there are some situations where you don't know what the problem is, and that's where we are sometimes a little weak, right? Is because our electrician walks in, the customer goes, "I got this problem," and they go, "Well, that's an electrical problem then." And, yeah, they could solve it, but it really may not have been the right way to solve it in some situations.
With AI, we can create some. Well, JP and I were talking about this earlier, some quite interesting new tools which start to sort of bring together those different capabilities so that you can actually create a diagnostic that is agnostic, if you like. It's not coming in as an electrician or a plumber. It's actually coming in in a way where you're saying: Okay, I can now look at that problem for you, and then I can actually interpret that problem in some really interesting ways, and then we can bring you in the people who are the right people to figure out and solve that problem. I think if we can start to do that kind of work, then actually we're in a brilliant place with customers because then we're still keeping the specialist capability.
We're not turning people into general-generalists, we're making AI be the generalist on our behalf, if you wanna call it that, and a wicked smart generalist, because it can actually... It can synthesize all of the really clever thinking, and experience, and knowledge of those individuals and put it into a bot. We were trying to come up with a name earlier for the bot. If we do, then, you know, that, that to me, is sort of, that kind of thing is the genesis of us doing some things that are quite transformative.
It's just apart from your hands. Maybe one for Peter first. If we look at Engine and you strip out Google, would that segment still be in decline for the first half, or do you think it would be actually up?
Google was off about GBP 2 million, so year-on-year, which was very much in a little bit of M&A, a little bit beyond. So I can't I think it probably still down a little bit, but less. I mean, interestingly, we had brands like Salesforce, Amazon, were actually up, so it was quite a sort of mixed performance in. But, you know, with Google being our biggest client, that was one of the reasons, you know, why it was in decline.
Okay. And the second one is just going back to AI. Obviously you're now doing the buyback, and there'll be limited, probably large M&A in the near term. So without M&A, internally, will you have enough investments to achieve your ambitions with AI? And then the second one, I guess, linked to that, is using and utilizing more AI, how does that impact your conversations on pricing with clients?
To be really clear on this share buyback piece, that slide that had the different uses of capital, they are competing uses of capital, and they are literally competing for that money. So the share buyback, we very deliberately said, "Up to," not "We are going to spend GBP 30 million," right? So if we need to spend the money on something else, buying a company, AI, whatever, we will spend the money on that. We are not going to say we're spending GBP 30 million, because we said we'd spend GBP 30 million.
If it's the right thing for us to do is, you know, if JP comes to me and says, "Right, we need to spend GBP 10 million on this, and here is the, you know, the case to do that," then, you know, Peter and I might sort of beat him to a pulp to make it into GBP 8 million or GBP 7 million or something, but, but we will spend the money on that, because if it's the right thing to do, we will, we will make that investment. If we have to turn around and say to shareholders, "Sorry, that GBP 30 million is now GBP 22 million," then I think shareholders hopefully will understand, especially if we give them the case, this is why we're gonna do that.
So, you know, I'm just... You know, that sort of, like, competing uses of capital phrase internally, is very much what we've been discussing, is this is a competition as to who can spend the money the best, and it's only if we can't find a really good use case for it, then we'll use it to buy shares in our own company. On the pricing piece, it's interesting. So far, procurement is not pushing the button on that, in large part because there is also a tremendous amount of fear within procurement connected to AI. They are mildly terrified about it destroying their own business. So they're mildly terrified about trying to renegotiate with people about something that could literally put their own organization out of a job.
And in large part because there's also still a whole massive amount of concern about legal frameworks around things. In other words, we are not at the point, at this stage, where we are auto-generating completely original content from AI that is being used literally within the market. In other words, we are not auto-generating email content out of a whole bunch of stuff that the client can't do a really compelling audit trail on. Because they are frightened to death about somebody saying, "You broke the copyright," or whatever it might be. Okay?
And until that starts to get solved, and we, I think, are still quite a way from solving that, what it's being used for more is prototyping, you know, that kind of work, where you're saying, "Right, okay, here is an instance of X." And we've had, in fact, some pitches where clients have said, "We used AI to generate, you know, a whole video set of content, a whole bunch of stuff," and the client goes, "That is awesome. Let's do it right now." And we're like, "I'm afraid that was generated with AI. We can't." And then they've had to go, "Oh, okay." But we can recreate it. It's just gonna have to be done in a still relatively traditional process, because... But what we have done is we've massively accelerated that process.
But that would have been a cost to us anyway, at a certain level. So at this point, it isn't changing the pricing. It will come, there is no doubt, but until a lot of the legal stuff really gets ironed out, I think we're still a little bit of a way off on that. Because it is this, like, do you have real first-party data? Do you believe it's yours? Have you completely checked it? One of the sad truths about first-party data is there is a whole load of first-party data that comes out of India that is not first-party data. It went to India as some other form of data and then came out of India. I'm horrible, I don't want anybody to think I'm being mean about India, but it unfortunately, it's like money laundering.
A lot of it goes in there and comes out. For a lot of old purposes, pre-AI, people were less... We wouldn't do it, but we know that there were other firms who would do that on areas like lead gen. They were using that kind of data, and we were like, "We can't do that. We don't. That's just too big of a risk for our organization." But smaller firms who weren't public, they would do that kind of thing. Our clients know this. They're not stupid, because they're like, "Okay, how did we get that first-party data so cheaply?" Sort of like, you know... It's kind of too cheap. It can't be. There must be something slightly odd about it. And in AI, that is a particularly worrying problem, right?
That is, you know, if you go talk to any of the big tech firms, any of the big consumer, you know, Unilever, Procter & Gamble, any of those guys, they are terrified of that. They're absolutely scared to death of it. So until that gets solved, I don't think the pricing thing is really a major concern.
So the technology vertical, which is 33% of-
Yeah.
The first half sales, can you give us roughly a like-for-like decline?
It was down to 14% in-
14%.
Yeah.
In like-for-like terms?
Yes.
Yeah. Okay.
Do you want to make a stab for the second half?
Less.
You're saying down. Are you saying down?
I think it'll still be down. I think partly because of Google, but I would guess it's probably low single digits down. Is my gut.
Yeah, we could, yeah, because, I mean, the comps last year, we were, you know, if the group was 31% up in H1, and I think we were about 10% in H2 or something. So, you know, we are entering sort of easier comps. And some of that is, you know, companies like Blueshirt and Palladium, who still had a reasonable H1 last time. So, you know, we are entering a world of sort of easier comps, which should make life a bit easier.
Okay, thanks. And then, on Mach49, you're making positive noises on that. Just in terms of the way that people are talking about the market, in terms of delaying projects and stuff like that, it sort of feels like Mach49 would probably be in that sort of zone. Do you agree with that, and does that make you worried?
Yeah, it has impacted them, for sure. And they are one of the ones who've got a big backlog of signed, like, "We're gonna do this. Can we start this?" You know, and whereas before they would have said, "We need to start tomorrow," now it's like, "Well, could we start this in three months time?" Type of thing. But it's starting. In fact, in the last, literally in the last couple of months, it feels like that has started to unlock. I wouldn't necessarily want to get too excited about that, but it's, you know, every time we've spoken to them recently, it's like, "Oh, yeah, this is now finally unlocked and is starting." They, to be perfectly honest, were also slightly guilty last year.
They were so crushingly busy that they just priced stuff in a way where it's almost like, "Well, okay, if you're stupid enough to pay that, we'll do it for you," type of thing. And, you know, when the heat kind of went out of that to some degree, they didn't change their pricing models, and so they've had to rethink some of the ways that their entry pricing works. But it is. It's fine. They've sort of unlocked that problem now, and it is in a good place. But yeah, the Kin + Carta ask for, sort of like, if it's a big number, "Can we delay it three months?" Procurement is brilliant at that.
Oh, yeah, sorry, I got one more thing that we need you to sign and check, and blah, blah, blah, blah, blah." That's what procurement gets paid for.
Thanks. And then, just notwithstanding what you said on tech spend in the second half, is it fair to assume that as the sort of cloud-based vendors launch commercial applications of generative AI, you know, like Copilot, whatever, and that's all coming towards the end of this year and next year, is that a proxy for your growth a bit, in terms of those customers are now launching those?
I would agree. I think this, it's an interesting challenge, right? Everybody's pricing it at sort of, seems to be almost everybody thinks $30 a head a month is the price. I'm not sure it is, and I suspect that price will come down pretty quickly. And that will therefore slow the revenue growth, which will slow the marketing spend. But it'll still be pretty good growth. You know, 'cause really money from Microsoft and people like that, good free money. You know, they bolt this thing on and see how many people buy it.
But from an activity perspective, is that like a sort of turning point for you in terms of, you know, these guys are then launching the commercial and whether people, you know, get their need to get the pricing right from beta to commercial?
Yeah. No, I know. I think it's the start of the next wave. Yes. I think it's the start of the next wave. How quickly that wave gets going is debatable, and I'm just trying to be cautious here. I'm just trying to not oversell it. I could say, "Yeah, it's gonna be amazing. It's gonna be, like, skyrocketing." I don't want to set that expectation, but I think it's gonna be good. I just don't know how quickly it's gonna really drive up their revenues. Mainly because they still generate crazy amounts of revenue from stuff that isn't AI. You know, so the proportion of the revenues that is gonna be impacted from an AI perspective is still gonna be a pretty small number to start with. It's just gonna be the fastest growing portion.
And that's part of our challenge, is we need to sort of- you know, like, we do work with OpenAI and a whole bunch of these people. We work with Google and blah, blah, blah, and all this stuff. So we can see what they're doing, we can see how they're growing, we can see how they're spending money. We just need to be closer to those parts of their organization. And if we are, I think we're in a good place.
Hi, yeah. Just, two quick ones from me. One is, be great. So is-- With Engine, it feels like you had a better result on the cost savings out of that sort of integration. I was wondering if you might be able to just give us a bit of color around the extent of that cost saving. The second one, actually, a bit broader and obviously think about capital allocation and the, that sort of priority exercise you have to go through on a continuous basis about where you're gonna be spending your money. Clearly, you've, you know, the share price has an impact on how high up share buybacks come in your thinking. Can you give us a sense of, you know, at what point does the share buyback become uneffective on economic perspective?
There was this whole debate about intrinsic value. And it's sort of a moment in time thing, right? You know, it's like, what is the company worth that day? You know, so is that GBP 9, GBP 10, GBP 11, GBP 12? What's that number? We just know that we're nowhere near it right now. I think it's the best answer that we can give. We know we're massively undervalued. I don't think you can put-- Yeah, if you're doing it today, realistically, you have to do kind of almost follow the same process that you guys follow on what do you think is the underlying value of the company? You know, do it several different ways, some of the parts, blah, blah, blah. You know, depending on which analyst you talk to, that's sort of-- there's a range there.
GBP 11-GBP 12 , sort of, isn't a million miles off in terms of the, sort of, what could the value of the company be. Do you discount that? I don't know. But the truth is, even if you say on the low end, it's something like GBP 9, and on a high end, it's GBP 12 or GBP 13. GBP 6, we're not bumping into that as a problem as to whether or not we should buy. I think once we start getting... You know, hopefully, we get into next year, and we start to see a recovery, as we, you know, we grow more, or we deliver better profit, whatever, then that's a nice problem to have, to try and figure out, okay, should we still be buying the shares or not? In other words, I'm dodging the answer.
I think on the engine synergies front, I think to me, you know, there's two areas we've added a lot of value. One, on sort of head office cost savings, where effectively they had, I think it was GBP 18 million. I mean, it was slightly confusing because it was all the property, and every finance person went to head office and, you know, some of the head office we actually put back into the underlying businesses. But I would say on cost savings, where we said we'd generate saving at least GBP 3 million, I'm gonna say at least GBP 4 million. And we have invested some of that in other areas, it's building, we've done a nice refurb, et cetera.
I think the other area in which I think we've done a fantastic job is on the margins, particularly for Transform. They had a slightly sort of bizarre model, whereby they win government work on the basis of cost and then staff with expensive freelancers. So it ends up being a sort of 5% margin business because of, you know, there was pressure on sort of pricing in terms of, you know, how they'd win business, but also their delivery model wasn't great. So they invested in technology, Kantar. They've got a very good CFO, who's really gone in and, you know, transformed the way they deliver work. You know, helped by them winning, you know, very good long-term contracts with the Department for Education. So that's now a 15% margin business.
So that in itself has probably delivered GBP 3 million extra profit, which wasn't there before. So it's probably a combination of, you know, efficiency savings, where we didn't need to head office and then really helping, you know, themselves improve their underlying profitability.
There's still more to go there, though. There's definitely more room.
Hi, Brady from Stifel. Coming back, apologies to AI, just sort of where you are today, which of your agencies would you say—and, and I know you've got many, so maybe just, you know, pick two. Do you see the most opportunity, and which one do you see the most potential risk from?
I think Savanta has a massive opportunity from an AI perspective, but it's because of the data that-- and their history and understanding how to use data, how to create the right pools of data, all of that stuff. So I think they are really, really, really well positioned. No guarantees they'll succeed and, you know, like Manchester City starts as favorite in the Premier League every year, but there's no guarantee they'll win it. I said that for JP's benefit 'cause he loves sport. Complete sports nerd. So I think they... Huge potential upside for them. I mean, that's partly why I think, you know, if you looked at the five-year forecast, that segment has probably the highest growth opportunity out of all of them.
And actually another business in that space, Planning, which is relatively small, it's a CDP business, it's a customer data platform business. Again, huge, huge, huge growth potential if we get that one right. In terms of the ones that have the potentially the biggest threat, I think it is the more traditional comms businesses. Because it's the thing at the moment where we generate quite a lot of content in those areas. If the legal stuff that I referenced earlier gets ironed out, then auto-generating content becomes really inexpensive, really fast. If we are able to build the bots that we think we can build, you will be able to turn pretty inexperienced, pretty low-cost consultants into mega consultants, and if we don't do it, somebody else will.
That's where there is a risk there of a race to the bottom in price, because it's seen as, it could be seen, if it's not done right, as a relative commodity business. It's already really a relatively commodity area. You just have to be really good at what you do, and you can do quite well, but that just sort of gets progressively harder and harder, and AI could just... If we're not-- if we don't do it right, AI just makes that problem bigger, doesn't make it smaller. I think, you know, this is then the challenge for the likes of MHP and M Booth and Archetype, these kinds of businesses is: how do we embrace AI and turn all of our people into, like, you know, take their superpower and AI enable it?
If we do it in the right way, we're in really good shape. I mean, it's like an analyst job, right? Analysts, if you simply just wanted to apply AI to this and just basically put you all out of a job, we could do it in about two months. You just train it on everything that you've ever written, and within two months, it would be producing better content than you can produce. Unfortunately, it's just the truth. We've tried it, and we know it can be done. That doesn't mean that you should be put out of a job. What it means is you should now be able to do more, better, produce richer levels, much more hyper-personalized versions. You should be able to produce an analyst report that's specific to BlackRock versus an analyst report that goes to everybody.
You should be able to produce a specific analyst report that's for Slater's, because they're very different to BlackRock. You know, so this is where AI creates opportunity, and it's a question of how you take advantage of that. But if you just, just say all you do is apply it, if all it does is efficiency, over time, it doesn't solve a fundamental problem, which is that the barrier to entry is not that high, you know, price point could get eroded, blah, blah, blah. So it then becomes a, "Right, how do I use this to make us produce a product that we couldn't produce before?" And that... For that area of our business, that's the challenge. It's not a, how do we use this product to produce what we're doing better? It, we should be doing that anyway.
But it's how can we do something that we couldn't have done before? That isn't easy, there's no doubt about it. You know, that we will fail as much as we succeed in that respect, I'm sure of it. You know, to some degree, if we fail, we fail. I mean, I'd rather fail than not try. You know, that's the way you learn. Thank you, everybody.