How to Get $500M to Build a Website
Blair tells the story of recreation.gov and how its performance pay deal has been nothing but wins for every party involved, with plenty of lessons for buyers and sellers of all kinds of services.
Links
"Performance Pay Leads to a $500M Website" by Blair Enns for winwithoutpitching.com
Transcript
David C. Baker: All right, Blair, today's topic is how to get paid $500 million to build a website. The entire crowd woke up. Oh, I want to learn that.
Blair Enns: This is not one of these spoof episodes.
David: Yes, right. It's sad that we have to tell people that anymore.
Blair: Because it's a bit of a spoofy title, isn't it?
David: It is. How to get paid $500 million to build a website. Who doesn't want to do that? Was this story in your most recent book, The Four Conversations?
Blair: No, I don't think so.
David: Wait, you don't think so? Who wrote the book?
Blair: You know how you write a book. When it's out and it's done, it's like the last thing you want to do is read that book. I'll read it in a couple of years. I'm sure it's very good because I didn't go deep into pricing, and I didn't get into performance pay at all. I have been sitting on this story for a little while. Then, when Booz Allen came up in the news in the last couple of months, I knew what the numbers were, what they had earned when I first heard of the story. It was in the range of $165 million. I went looking for an update, and my mind was blown.
David: I'll just be honest, between you and me and thousands of people, I wasn't sure this was real either. I knew it was real, but I didn't think it was this real. I started looking at it, and it's like, "Shoot, this is one of those times that Blair is not exaggerating." Tell us the story. You and Colette actually experienced using this thing, too.
Blair: When I knew I had to use the app, oh, I was so excited to use it, and it was everything I hoped it would be. I've been aware of this story for about three, maybe four years. This is 2025. Yes, so around 2022. It was the front page story on The Wall Street Journal above the fold, the large type. It was actually not a negative headline. It was, Americans are flooding back to national parks, and one consulting firm is reaping the rewards.
Let me leap forward in time. Some of the listeners might not remember Doge, but it was a big thing back in May of 2025. At peak Doge, when Elon and his team were going through the US government finances, basically looking for fraud, waste, and abuse, one of the things that they uncovered was this consulting company called Booz Allen Hamilton, goes by BA or Booz Allen, 12 billion in annual revenue, when the source of 98% of the revenue was the federal government.
David: That might be a client concentration issue. I don't know. Seems like it qualifies.
Blair: It's spread out throughout numerous departments within the federal government. In a climate where people are looking for government waste, these numbers, 12 billion and 98% of it coming from the federal government, just sent some people into this stratospheric-- They just lost their shit over it. I knew at least where some of that revenue came from. I knew some of it came from the website that is known as recreation.gov, which Booz Allen won in somewhere around 2017. The website launched in 2019 or so.
I knew that hundreds of millions of dollars of that 12 billion came from this contract. I thought it was a couple hundred million. Turns out it's approaching $500 million, and it will be somewhere in the neighborhood of $580 million by the time their 10-year contract runs out. I thought they didn't deserve the flak they were getting because I was so proud of this consulting company that I don't know, and I was so proud of the government procurement people who hired them on this basis. This firm got rich- within the scope of their revenues, they're not getting rich. They earned a massive amount of money, and it cost the US taxpayer absolutely nothing. The media went crazy over it.
David: They thought it was crazy that Booz was making that much money?
Blair: Yes. Anytime you see a government contractor making hundreds of millions of dollars, your immediate thought is who screwed up or who conspired on the government side with their cronies at the consulting company to feather their pockets to the tune of hundreds of millions of dollars. That's not how it went down. Now, if I back up just a little bit, I'll do a comparison of two different US government websites. Last March, Colette and I are down in California. We want to go hiking in Joshua Tree National Park. As we're pulling up to the park, we realize, "Oh, we need a pass." The sign says go to their website. You go to their website, you're redirected to recreation.gov.
I thought, "Oh, shit." I've been to the website, but I've never used it before. "This is a government website, and I have to make a purchase. How onerous is this going to be?" I know it's going to be a good experience because I know Booz Allen just killed it on this project. It's this beautiful user experience, very intuitive. Within two or three clicks, I'm paying for my park pass with Apple Pay on my phone. Then we pull up to the gate, show the pass, and in we go. It was just like going into Disneyland. I've never been to Disneyland, but it was just like going into a really well-run private enterprise. It was very ungovernment-like.
Now, if I contrast that experience with another experience, as soon as I got home, I had to start working on renewing my-- I'm Canadian. I have a US visa. I've had one for going on 20 years now. I have to renew it every five years. I had to renew my US visa. Now, the websites involved in this process--
David: A little different.
Blair: It's everything you would expect from an online government experience. This isn't a knock against the US government or the federal government. You would expect this type of experience with almost every level of government in any country in the world, except for Estonia or some of the really technically advanced countries. Nothing works. It crashes all the time. The wayfinding is horrible. It contradicts itself. It looks dated. It feels dated. Doesn't work on some browsers.
I would say it's a frustrating experience, but it's not because we know going in, this is an online government experience. It's not going to be easy. The onus is on me to do all the work to figure it out, and there just seems to be no impetus on the part of the provider here to make it easy. That's just because of the bureaucratic nature of government. Anyway, I'm not trying to call out the US State Department or whatever the website is. I'm just saying that's the norm, and we all accept it.
Recreation.gov is an entirely different experience. This is like a private enterprise. Everything from the design, the entire user experience, the way you can pay on your phone, they give up the whatever percent to Apple. It's just seamless. It's like this was not designed and built by a government agency. It wasn't. It was designed and built by Booz Allen Hamilton. The reason this web experience is so freaking good is the way they were paid.
David: Right. Not cost plus. It's not like we don't yet have Air Force One after 4,800 years. I have my own experience with this. Last night, I was filling out-- I'm starting to do some work at a maximum security prison here in Nashville.
Blair: On the inside or the outside?
David: Hopefully, on the outside. I'm going to be on the inside, but I have to get all this paperwork done so that I'm not going to fly out everybody in a helicopter or something.
Blair: I'm sorry. What are you doing in a maximum security prison?
David: Oh, I'm working with the prisoners about life skills and so on. I have to be approved by the state level to do it. It's a maximum security, so they're a little careful about it, understandably.
Blair: Is this a whole other podcast, or do we need to talk about this now?
David: No, we don't need to talk about it now. You're distracting me. My point was they're asking, there's a home phone and a work phone, and there's no contact for mobile phone. How many years has it been since I've even had a home phone? Anyway, yes, I understand this completely.
Blair: Yes, that's like the experience of a typical government website.
David: Exactly.
Blair: It asks you for this information. What's your fax number? You can't proceed until you give them a fax number.
David: You make one up, and nobody will ever check on it. You went in. You were expecting a good experience, but it was even better than a good experience. Your point is that the reason this was done was because somebody decided to price this differently to arrange the incentives differently, right?
Blair: Yes. I don't know anybody on the client side of the US government. I don't know anybody at Booz Allen, but I've read quite a bit about it. I had ChatGPT 3.0 do a whole lot of research, and then I went down some rabbit holes to read some of the backup documentation. It's a fascinating story. I wish I knew a little bit more about it. Here's the deal. In 2017, the US government puts out this tender to basically unify the e-commerce transactions of 14 different departments. It's all the outdoor and recreation stuff, so national parks and a bunch of others.
They get a bunch of bids from the usual large consulting companies. The Booz Allen bid basically says, "Listen, we're not going to charge you anything for this. We're going to incur all of the upfront costs, so all of the development costs, all the research costs. We will build the website and maintain it. We will build and maintain the mobile app. We will build and run a call center to handle all of the phone calls that arise from it. We'll basically take care of all of the customer interaction. It doesn't cost you anything. We're going to add a small transaction fee." I think it's about $6. Every time somebody buys a park pass or a lottery ticket to an exclusive event or a camping site, or the various things that people might buy, typically parks-based. "Anytime somebody buys something, we just add a small fee onto the transaction, and we'll take that fee."
They won the bid, and in the bid, from the historical information that the client, the US government, had provided, they estimated openly that they would earn about $185 million in top-line revenue over the 10 years of the contract. It was a five-year award and then five single years of renewal if they hit certain benchmarks of performance and customer satisfaction.
David: They would have accepted it at that level of projection? They still would have made money?
Blair: Yes, they still would have made money. I have no idea what their cost would have been. They would have been in the millions of dollars, but obviously, if they had $185 million, that's a significant amount of revenue. I think, in the back of their minds, they thought, "Well, this is based on historical, and the historical user experience has been horrible, and we can probably improve it quite a bit." Boy, did they improve it quite a bit. The first full year it's active is 2019. The bid process is in 2017. They're building it out through 2017, 2018.
It goes live somewhere around 2019, where they have 3.8 million reservations, so all of the activity, things that people can buy on this app, they have reservations of some form. Then in 2020, it goes to 4.5 million. Now, this is COVID. A, they did a good job. B, they got lucky with COVID. 3.8 million, 4.5 million in 2020. 2021, people start traveling. A little pent-up demand. 7 million reservations. 2022, 10 million. Then it drops a little bit, 9.5 million in 2023, 11 million in 2024, 12 million projected in 2025. They're projected to hit 13.5 million by 2028. 10 million more reservations a year than before they took over the project in 2018, 2019. They are going to earn $580 million from this project.
David: I did a little research after seeing your topic, but I didn't see this. Who was driving this on the Booz Allen side? Maybe not the specific person, but what was it that prompted them to think about this approach? What was driving it? That's the interesting part because this is not happening that much in our industry. I don't know if that's a supply issue or a demand issue. You point out later that there are some things about this story that make it hard to replicate, like we're not as big as they are, we don't have the capitalization, so we can wait that long to get paid, and so on. This is a big philosophical question. Is this not happening more because of the supply side or the demand side?
Blair: It's a great question. I've already done way more research on this post than I typically do on any post. I don't do research at all. If I were a proper journalist, I would pull on this a little bit more because I think there's a fascinating story here. I would like to know the inside story of the conversations that went on at Booz Allen. I would like to know who the brave procurement people are who approved this on the client side.
I've told this story before. I once said to a friend of mine who's a procurement professional, and we weren't friends at the time. It was the first conversation that we'd ever had, and it was on the subject of procurement. I said to him, "I want you, the client, to say to me, I'll make sure that if you make us a whole bunch of money, you'll make money too," or something to that effect. He said, "Blair, I'd get fired for saying that."
That's effectively what happened here. The buyer of this, whoever accepted this tender, looked at this and went, "Okay, if you make us a whole bunch of money, you're going to make a whole bunch of money." Then, on government, you have to contend with why this person is so brave. It seems like a no-brainer because the consulting company took all the risk away from the client, except one form of risk, and that is the risk of public exposure.
In government procurement, the last thing you want to read about or hear in the news is any project that you were involved in. This made the front page of The Wall Street Journal. It would not have surprised me if the contract had been cancelled, if people had been fired. I guess it blew over. The renewals kicked in. Booz Allen is still delivering on the service and they're adding other features as they go. There are other articles written on this. Because of the way they're paid, they're paid based on performance, they are going to the client with suggestions on, "Hey, we can add this." There's no additional cost to the client. They're investing their own money, and both parties are making more money. I just think it's bravery all around. It's brave for Booz Allen to bid this way, and it's even braver for the U.S. government to purchase this way.
David: I started poking around to see what the basic sentiment was around this. I was surprised in that I read over and over again that, yes, that's a great example, but you need to be really careful because for every one of those, there's many examples of where this failed. Then the instance that most of these sources cited was the, I don't know if I'm pronouncing this correctly, but it's called the FiReControl Project in the UK, where they were supposed to unite something like two dozen call centers into nine, and there would be better uptime and everything. The risk to the buyer in this case was non-performance. It didn't work. It was a disaster. Everybody was fired.
Then, looking deeper, there were two reasons why this one failed, and the Booz Allen one did not. All the sources agreed on this. They said that one of the reasons was because the contract wasn't written well. I presume by that they mean that Booz Allen had to hit certain things. There were net promoter scores and so on. The second thing is that there was no buy-in from the stakeholders. You had people who were dragging their feet and not cooperating helpfully, and so it fell apart.
The first point, I understand that one. The second point, it makes me really curious about how somebody on the government side built the system so that they got buy-in and didn't drag their feet. That would be interesting to me, too, if we ever figure that out.
Blair: Yes, that's a good point. I have a list of four reasons that I can cite for why this project worked this way. I didn't think about that last one. If I go through the list, number one, if you're a web dev firm or in the digital transformation business, and you're trying this on for size, this type of performance pay project isn't going to work with everybody.
The first reason it worked is the customers were already there. You have an established client with an established customer base. This wasn't a startup with a large total addressable market, but no actual or few actual customers. Because there were years of history of people purchasing park passes, et cetera, it was easy enough for the consultant to model out what revenue was likely to be. In their bid, they shared what the revenue projections were, $185 million on the top side. The customers were there, they could predict revenue, and then they could probably make some sort of prediction of their ability to impact revenue. The second reason it worked, I think, is the costs were shifted from the client to the end user.
David: Still pretty nominally. It's only $6, so it's not going to break anybody's bank.
Blair: Yes. Let's say if you look at comparative websites, let's put this in the $100 million range. I don't know, maybe it's $10 million, maybe it's $200 million. If we call it $100 million that the government would have to pay for this, they end up paying nothing, and that $100 million gets born by the end user in a nominal $6 charge. The markup on the transactions are nominal.
Another point here, under the idea of shifting costs from the client to the end user, the client was never upside down revenue-wise. They didn't have to finance any of this. Just imagine the procurement person does not have to put forward a proposal to spend tens of millions of dollars. They don't have to spend anything.
David: That's pretty powerful.
Blair: The third point is the consultant took all the risk. There's risk that they get paid zero. They invest, let's say, tens of millions of dollars, they get paid zero. That's unlikely. They would have to really screw up to get paid zero. It could be a poorly profitable job, or it could even be unprofitable. Then they're on the hook for five years. They're taking all the risk away from the client. Then, as I wrote that, I thought, "Well, there are always edge risks." In this case, one of them actually came to fruition, and that is a global pandemic. In the end, I think the global pandemic actually helped their cause, but probably at the time it was happening, they went, "Oh, shit."
David: Nobody's going to be traveling. Then they realized, "Oh, well, they'll be outside where there's less risk of contraction, so maybe this'll catch up."
Blair: Maybe, in my last point of why it worked, this might speak to it. Maybe there's a force majeure clause in the contract or something, "If there's a global pandemic and people no longer go to national parks, you pay us X," or something like that. My last point was one that you already made when speaking to the UK example. It was all tightly contracted. It's easy to imagine the politician screaming, reading the story in The Wall Street Journal, and wanting to be a hero and screaming, "I don't care if there's a contract on the renewal, make sure that $6 transaction fee gets dropped to $0.50," or something like that.
I don't know this for a fact, but it seems pretty obvious to me that would have been a weak spot in the contract. The contracts, I think, they are renewed automatically because their performance numbers and the NPS numbers were hit. The renewals are automatic. There was no way for somebody, typically it would be somebody in procurement, in this case it would be a politician, to get in there and try to be the hero by robbing the consultant of the money that they've earned.
David: Part of that probably goes back to your second point about passing the cost along to the user. If this money was not coming from people who actually enjoyed the park, then there would be people screaming, "This is not a good use of government money." This money is coming from the people enjoying the park. We're not asking somebody who doesn't go to the park to subsidize Booz Allen. That would have had some impact, too. It's a little harder to fight that, I would think.
Blair: To be clear, though, there still were some people screaming and trying to claw this money back because they just see the headline. Counterfactuals. I was looking for one, thinking, "Okay, well, typically, government buys big things on a time and materials basis." What's a counterfactual? What's an example of, if this had been sold on a time and materials basis, what might it have looked like? I think if you look at healthcare.gov, probably a more complex project and website, but probably in the ballpark of this.
I was trolling for a US federal government website build disaster, and I didn't have to look far to find healthcare.gov. Here are the numbers. September 2011, consulting firm CGI Federal, which is a Canadian company, by the way.
David: That's good to hear. We're not taking all the flak here.
Blair: The parent company, CGI, is Canadian. This is a US subsidiary.
David: I'm tired of you shitting all over my government.
Blair: I have to say, I want to be clear, I am not shitting all over your government. My visa has been renewed, and I want to give my thanks to the US government and the US people for letting me do business in your country. I do not take it for granted.
David: All right. Kiss up over. Back to the--
Blair: 2011, CGI Federal is awarded the contract to build healthcare.gov, which is their consolidation of the healthcare marketplace websites. They get it on a, "time and materials fixed fee basis." The contract allows for the reimbursement of all allowable costs. There are all these descriptions of what are allowable costs. The allowable costs are capped at $56 million in year one and a total ceiling price of $94 million after five years.
The US government is saying, "Okay, we will pay up to $94 million over five years to develop this website on a time and materials basis." They know if they hit the cap, the vendor, the provider, the supplier is going to bill to that cap. They're not going to bill a dollar less. They know it's $94 million. This is a four-year project. Three years in, they've billed $209 million.
David: That's higher than 94, isn't it? Last I checked?
Blair: I don't know how this works. The scope creeps, the allowable costs get amended. They're three years in, they're at $209 million. The site launches and crashes on launch and was deemed an absolute failure. The whole thing was largely scrapped. The government then releases CGI Federal from the contract. They hire another consulting company, Accenture, to come in and do the rescue work. Accenture does the rescue work and then takes over a new contract to manage the site over five years. Accenture earns $563 million in time and materials.
The same amount that Booz Allen is going to earn on a full risk performance pay basis over 10 years, they earn it in time and materials over five years. Then there's the $209 million that was wasted on CGI Federal. Then the government accounting office, you'll love this, estimated the total spent on project had hit $1.7 billion by 2014. You think Booz Allen is taking advantage of somebody? No. It's win, win, win. They win because they're helping the US government win. They've generated, what is it? $3 billion in revenue for the US government. From that, they took $580 million in fees. They win. Their client wins. The end user wins. Journalists win because they get to write these horrific stories, bending the truth when the reality is this is nothing but wins all the way down.
David: Podcast hosts win because they tell the truth.
Blair: I think the US government should let Booz Allen run all their websites.
David: Did you realize that you started to spell Booz with an E at the end halfway through your notes to me?
Blair: Is there no E at the end of Booz?
David: No.
Blair: I'm quite familiar with Booz, and it has an E at the end of it.
David: Oh, maybe you're right.
Blair: No, you're right. Yes.
David: Can I add something here? This got me thinking about the whole legal profession, too, and how much they make in fee billings per full-time equivalent. As far as I'm aware, the only branch that is 100% they take all the risk is the PI side, personal injury side. I thought, "Okay, how does this show up?" Their average fees are $243,000 per full-time equivalent employee. The rest of the industry varies between $130,000 and $175,000.
They make a whole lot more money, which didn't surprise me all that much, but this was the kicker for me. They do that on the worst freaking advertising in the world because then I started to think, "Okay, well, if I were on the supply side and I wanted to do more of this, what kinds of projects would I be interested in taking a look at under this model?" I'm thinking hypothetically here, I'm getting clients coming to me. I'm an agency, and I don't really think this is necessary. It's not all effective. There's no way I would take the risk on something like this.
It would change the kind of work you did as well. This is a point you made in your pricing book. If you were only going to be paid on performance, you would be a lot pickier about your clients, too.
Blair: Yes, very picky. I had a conversation yesterday. A friend is in town. He was talking about he was bidding on a large project, and he wanted to bid on the performance pay because he had read my pricing book, Pricing Creativity, a Guide to Profit Beyond the Billable Hour. He was putting the bid together with a veteran, a person who had deep experience in selling large contracts of this type to these types of organizations.
The guy said, "All your motivations, everything you want to do, it's all correct, but here's what's going to happen. When we go ask to get paid for the incentives that we've earned based on hitting the KPIs that we've agreed to, somebody in the organization is going to say, 'If I write that check, I'm going to get fired. Therefore, I'm not paying you." He ended up following his colleague's advice, bidding it out, I don't know if it was time and materials or what. They took the incentives out of it.
I said, "You ended up doing the right thing because performance pay is a form of partnership. It's when your financial incentives are aligned, and you would not partner with just anybody. You have to have this deep trust in your partner." Now, did Booz Allen have deep trust in the US government? Probably not. The challenge there is there are so many people that could scuttle this. I think this is where a large organization with a good bank of lawyers and a solid contracting process, that's where it makes sense.
Generally speaking, I actually think performance pay pricing makes sense with your clients who are founder-led organizations. If you have that trust, founder-to-founder trust in the two organizations, then you can proceed with contracts and trust in the contracts as a fail-safe. At the end of the day, you really have to trust the person you're in business with.
David: You're the person that needs to draw the right conclusions here, but if you ask me, what do I want people to take away from this, I would say you need to start looking for an opportunity to experiment with this. Even if nobody accepts it, if you leave it as an option in your pricing matrix, it will demonstrate so much more confidence to the buyer. That's what I would walk away with. What would you want them to walk away with?
Blair: I think that's a great summary. Your last point there, I don't know if we've made this point ever, or certainly not enough in the conversations that we've had. Just by putting forward a performance pay option as an anchor price in a multi-option proposal, that signals to the client so much confidence. It signals to the client your confidence in your ability to actually create value for your client, and it increases the amount you will be able to command in the other options. I think that's a great takeaway lesson.
David: Great. Thank you, Blair.
Blair: Thanks, David.