New here? My name is Ali and I write about the art of building digital products, transforming your business with AI, and outsourced IT services. Subscribe if these subjects interest you. Ooo also if you’re interested in b2b businesses join my community.
I started Wednesday in 2019. If you don’t know Wednesday at the time was in the business of outsourced product engineering. That’s a fancy way of saying, we helped onshore digital businesses with product strategy, design, and engineering at an affordable rate.
Thousands of firms in India alone offered the exact same services. Globally the number is far bigger, and it gets more crowded every month.
Over the years in this business I’ve built a small circle of friends. Some have exited, some are still building, each at a different phase. What follows comes from my own experience, checked against theirs. So here goes.
Services scale worse than software per unit of capital. It’s a slow business. People based, referral based, hard to grow. Every services founder has heard this, and most of us have said it. Hell I have said this multiple times. I no longer believe it.
Indian IT services was built on headcount. Revenue was engineers multiplied by a billing rate, so growth meant hiring, and hiring was the hard part. We faced this ourselves. Finding good people and getting work out the door was the bottleneck for years.
That bottleneck is gone. Wednesday, for example, now runs a fleet of agents per project and delivers the same customer experience with a third of the people a project used to need. Enough has been written about what an AI-native services business looks like, so I won’t repeat it. What I want to write about is growth, because removing the constraint changes what you can aim for.
A venture-backed startup is expected to double (or more) every year. I think an IT services firm can now grow at that rate. The rest of this piece is about why most won’t, and what to do about it.
Elements of demand & supply
Every services firm has a supply side and a demand side, and growth depends on the two matching. So it’s worth being precise about what each one is made of.
Supply is everything your firm puts out into the world. It includes:
Your website, which says what you do and for whom
Case studies, which show what you’ve done before
Testimonials, which say someone was happy with it
Services and offers, which say how you package the work and what it costs
Socials and the founder’s LinkedIn, which show what you think about
Each of these is a signal to a buyer about what they’d get if they hired you. Put together, they are the shape of your firm as seen from the outside.
Demand is simpler. It’s a to-do on the buyer’s list. It’s near the top, they cannot avoid doing it, and they’ve tried the obvious options and none have worked.
It looks like this:
They hired a chief AI officer a year ago. Operating cost hasn’t moved and nobody can point to a process that runs differently. The internal team has no experience with this, every POC has stalled, and the board wants results in 2 months. They’re now looking for someone who has done an AI transformation before.
They ran paid ads for six months and got leads the sales team couldn’t close. The founder has a following on LinkedIn but no system behind it. They need pipeline this quarter, and they’re looking for someone who can turn the founder’s brand into a lead channel.
They need twenty engineers by March for a contract they’ve already signed. Their two staffing vendors have sent them forty CVs and no one worth hiring. They’re looking for a team that can start in weeks.
Their core platform was built by a vendor who no longer picks up the phone. An audit is due in six months and the stack won’t pass. Their team can keep it running but can’t rebuild it. They’re looking for someone who can get them compliant without a rewrite.
Every one of these has three things: a date, something that already failed, and a gap the buyer can’t close alone. When all three are present, the buyer is looking for you. When any one is missing, they aren’t, and nothing you put on your website will change that.
Growth
Growth happens when your supply is the shape of somebody’s demand. Think of it as a puzzle piece. The demand has a shape. Your job is to find it and cut your supply to match. Most firms do it the other way round. They cut a piece, then go looking for a hole.
This is what it looked like for Wednesday before AI. A founder has raised a seed round and has twelve months of runway. The investors want a product in market by month six. The founder has tried hiring and found that senior engineers won’t join a company with no product, and the two freelancers she found have built a prototype that can’t be extended without starting again. That’s the demand: a date, a failed option, and a gap she can’t close herself.
Our supply at the time was: “We take you from idea to MVP in twelve weeks with a team that stays on for the next iteration.” Same shape. She didn’t need us to explain what we did. It was easy to sell and fulfil the todo.
That match is what growth is made of.
Can services have exponential growth?
Yes. The constraint on a services firm was always people. Every new project needed new hires, and hiring was slow, expensive, and hard to get right. That put a ceiling on how fast you could grow no matter how much demand you found. The ceiling is gone. A team of three with agents can deliver what a team of ten did two years ago, and it can take on the next project without waiting for the next hire.
And yet the average digital agency grew 7.5% last year. Nobody is doubling. A business that has lost its bottleneck is growing at the same pace as when it had one.
The reason is where the money goes. When a services firm decides to grow, it spends on supply. A new website, a new deck, a rebrand, a video for the case studies page, a fresh set of service lines. I did all of this. What almost nobody does is go and find out what demand looks like right now, who has a to-do, what they’ve already tried, why it failed. Which is a strange choice, because supply is the part you control and demand is the part that decides whether you get hired.
Fixing this is a mindset change more than anything else. Six things have to move, and I'll take them one at a time.
1. Rate of growth matters more than growth
Most services founders plan for growth. Twenty percent this year, thirty if a big client lands. That was a sane target when every new project meant new hires. It’s a low one now.
Rate of growth is a different question: how fast can this compound, and what would have to be true for it to double. A venture-backed startup is judged on that rate and plans backwards from it. Services founders have never had to ask the question, because the answer was always headcount. It isn’t anymore. Set the target at 100% and work out what has to change to get there. The target moves first. The strategy follows it.
2. Competition matters less than you think
Almost every services founder I’ve spoken to, from the two-partner shop to the firm with a thousand people, makes the big decisions the same way. Which service lines to add, which markets to target, what offers to run, what kind of customer to go after. They look at what competitors are doing and do a version of that.
Demand for the category was already proven, and the constraint was supply. Whatever you did, you were going to grow ten or twenty percent, because that’s how many people you could hire and bill. So the sensible move was to copy a working model, do the work well, and grow a little faster than the market. That’s the whole strategy, and for two decades it was enough.
The bottleneck has moved. Supply is no longer the thing that caps you, demand is. Consider what that means by looking at a product company. If it wants venture backing, it has to grow fast, and there is only one way a product company grows fast: it finds a to-do that buyers are desperate to get done, that they cannot put off, and that nobody has a good answer for yet. Everything else, the roadmap, the pricing, the marketing, comes after that. The to-do comes first, because without it there’s nothing to grow.
Services firms have never worked this way, and most have never needed to. Finding an unmet to-do is a muscle the industry never built, and the Indian firms I’ve seen are the weakest at it, because the model here was cost arbitrage on proven demand for so long. But the reason the muscle didn’t matter is gone. You are no longer supply-constrained, which means the thing that let product companies grow fast is now available to you.
The fix is simple to state: stop looking at competitors at all. The only thing worth studying is the buyer’s todo list.
3. Find the new to-do
Buyers’ lists change. Two years ago few small firms had “run the operation with fewer handoffs” or “answer every customer in under a minute” near the top of theirs. Now they do. Upwork’s Q1 2026 survey of SMB leaders (10 to 99 employees) found efficiency is the most anticipated outcome from AI agents over the next two years, cited by 68%, and customer service is where the most pilots are running, at 40%. One in three call AI agents mission-critical to company strategy.
What makes this demand, is that the first attempts haven’t worked. 74% say AI has improved productivity, but for most the gain is under 25%. The to-do is at the top of the list, the pilots have run, and the result is a modest gain and no clear return.
In another two years the to-do will be something none of us can name yet. Fast growth comes from spotting it while it’s still new, before it has a category name, before there’s a rate card, before there’s a competitor doing it whose website you could copy.
4. Sell before you build supply
When a services firm decides to reposition, it does the responsible thing. It runs user research, reviews the competition, writes a new ICP, defines the pain points, builds an offer. Six weeks later there’s a document. The document becomes a new website, a new deck, a new set of case study videos. Then the firm waits for the phone to ring.
I’ve done this. It produces excellent material and no growth, because every one of those artefacts was built from what the firm thought demand looked like, not from what it heard a buyer say.
There’s a faster way to find out, and it’s cheaper. SELL. I’m not asking you to hire a salesperson. I’m talking to you the founder or someone in the executive team. Get on calls with people you think have the to-do and try to sell them the outcome. Every call tells you whether the to-do is real, whether it’s urgent, and what they’ve already tried. Ten calls teach you more about demand than a research project does, and you might close one.
If you’re stuck and want the mechanics, here they are. I’ll write a separate post on each.
Find the to-do, or guess it. If you have a year of sales calls, mine them. Listen for the same sentence coming from different people, and write it down in this shape: “We need to ______ by ______. We’ve already tried ______ and it didn’t work because ______. We can’t do it ourselves because ______.” If you don’t have calls, you don’t get to skip this step. Fill the blanks yourself, as a hypothesis, and go sell it. The sprint will tell you whether you guessed right. Either way, the last two blanks are what you’re selling.
Run a sales sprint. Pick a cohort of twenty people who should have that to-do, reach out, and try to sell them the outcome. Do it every day for a week. Then change the cohort, or change the to-do, and run it again. The point is speed of learning, not close rate.
Look for pull. After each sprint, listen to the calls. You’re looking for the person who didn’t need convincing. They had the to-do, they’d already tried something, and they moved the conversation toward a start date without you pushing. If you had to explain what you do more than once, it’s a red flag.
Then scale. Once one cohort is pulling, and only then, put the supply behind it. Website, deck, paid, webinars, events. You now know what they say, so you know what to write.
5. Say no to referrals
This one is hard, and it’s hardest for the firms that are doing fine. If you have customers, a team to keep busy, and a reputation that brings work to your door, saying no to that work feels wrong. Every instinct you’ve built over ten years says take it.
Most services firms run on exactly this. Haus Advisors’ 2025 benchmark of development agencies found 82% rely on referrals and word of mouth as their primary source of new business. In a supply-constrained world that’s the whole model: the network produces a project, the firm staffs it, the client refers the next one.
6. Be something to someone
Open the website of any large services firm and try to answer two questions: what do they do, and for whom. You won’t be able to. They do everything, for everyone.
This is what happens when a firm grows on referrals for long enough. Each new client adds a service line. The result is a menu. A buyer with a burning to-do doesn’t want a menu. They want the one firm that has done their exact thing before.
The data on this is unusually clear. Promethean’s 2026 agency benchmark found agencies that reduced their services grew fastest, at 13%, while those that expanded grew 9.8% and those that held steady 5.7%. The firms that narrowed also posted 30% net margins against an industry average of 13%. Cutting what you offer made firms grow faster and earn more.
So be something to someone. Pick the to-do that pulled in your sprints, and let that be what the website says, and what the case studies show.
Where this leaves you
I wrote this for the founder who is sitting on a firm that works, that grows a bit every year, and who has a feeling it could be a very different business now. You’re right, it can. The thing that stopped services firms from growing fast was hiring, and hiring isn’t the bottleneck anymore.
So set the target at 100%. Heck a 1000%. Change your mindset.