Your biggest competitor will never show up in your pipeline

Melissa Kwan's customers are leaving. When they say why, they give the same reason: AI.

What makes it sting is that she has seen this story before, from the other side. Not long ago, her company, eWebinar, walked away from HubSpot. One person, working part-time with AI, rebuilt what they needed in five months.

Her verdict: "What we did to HubSpot is what someone out there is doing to us."

Someone out there is probably doing it to your company too. You just won't find them on your battlecard. I know, because I'm one of them.

The deal nobody knew they lost

Running Ocean Wave Growth on my own, I meant to buy two tools: a CRM and an email campaign platform. That's what a proper consultancy has, right?

I didn't buy either. I track my prospects in a Google Sheet and use ChatGPT, which I already pay for, to write and automate my outreach emails. It's not elegant. But it solves the problems I actually have, without adding the complexity I don't need yet.

Here's the part that should worry SaaS companies. No CRM vendor knows they lost me. I never booked a demo, never started a trial, never churned. I'm not on anyone's lost-deal report. I simply never showed up.

That's the new competitor. It doesn't appear in your pipeline, so it doesn't appear on your battlecard either.

Your battlecard is counting passes

Picture your SaaS competitive analysis. Five vendors down the side. A feature grid. A pricing column. Objection handling for each. Perfectly normal, thank you very much.

There's a famous psychology experiment where people watch a video of a team passing a basketball and count the passes. Halfway through, someone in a gorilla suit walks into the middle of the scene, beats their chest and walks off. About half the viewers never see it. They were too busy counting.

Most SaaS leadership teams are counting passes. They watch the vendors they already know, feature by feature, price by price. Meanwhile two rows are missing from the battlecard, and they're walking straight through the middle of the market.

Squeezed from the top and the bottom

Melissa's story is the pressure from below: smaller customers building "good enough" versions themselves. At the top of the market, something different is happening.

Jason Lemkin at SaaStr recently laid out the contrast. Sierra, which sells AI agents that handle customer service, reports adding $150M in annual recurring revenue in a single quarter, going from $200M to $400M in five months. Meanwhile, ServiceNow's shares fell about 27% in one day, even though its subscription revenue grew 21%.

Lemkin's rough maths on Gartner's forecasts makes the point: total software spending up about 15% in 2026, agentic AI up 141%. The budget isn't growing to make room for AI. It's being re-split. His advice to SaaS companies is to "plan as if no new budget is coming", and to be able to name the line item you replace.

hese are an investor's reading of company-reported figures, not audited truth. But the direction is hard to ignore. Enterprise buyers are moving money from tools people use to agents that do the work.

This isn't new. AI just made it impossible to ignore.

When I built battlecards in previous roles, the biggest competitors were rarely other vendors. Most of the time they were "do nothing" and "we'll build it in-house".

At an infrastructure software company I worked at, many open-source users never became paying customers. They had built their own tooling around the free version, and some had built their careers on it. Our paid product was more secure than what they'd put together. It didn't matter. Moving to it meant giving up something they had made with their own hands.

Psychologists call this the IKEA effect: we value what we build ourselves more than what we buy ready-made. Add the effort of changing a habit, and a home-made tool can beat a better product every time.

At a contact centre software company, the real rival wasn't another chatbot. Our chatbot was good. The rival was the status quo: people on the other end of the phone. Companies that saw customer service as part of their promise were afraid that handing queries to a bot would cost them the personal touch their customers counted on.

The lesson I took from both: people don't buy tools. They buy a solution to a problem. If you only compare yourself with companies that make similar tools, you miss the biggest part of the market. Those buyers aren't thinking about tools at all. They're thinking about their problem, and about which way to solve it.

In an earlier post on the 25-word test, I argued that the category you name decides who you compete against. The same goes for the alternative you name. AI hasn't invented the "build it ourselves" competitor. It has just made it cheaper, faster and far more common.

The two rows missing from your SaaS competitive analysis

So here are the two rows I'd add to every SaaS battlecard. Give them names, because what you name, your team will start to see.

The Agent. The buyer doesn't buy a better tool for their team. They buy an AI agent that does the job instead, and pay for outcomes rather than seats. This is Sierra going after the customer service budget.

The Weekend Build. The buyer, or one capable person on their team, puts together a "good enough" version with AI, a spreadsheet and a few automations. This is eWebinar leaving HubSpot. It's also me and my Google Sheet.

Once these two have names, they stop being vague market trends. They become competitors with strengths, weaknesses and a real cost, which means you can position against them.

What I'm still happy to pay for

There's one tool I wouldn't replace: Buffer. It schedules my LinkedIn posts for the week, and it costs me $6 a month for one channel.

Could I build my own connection between Claude and LinkedIn? Probably. But then I'd own it. Every 60 days LinkedIn would make me sign in again, and every time it retires an old version of its API, I'd be the one fixing the connection. When it breaks on a Monday morning, that's my problem too. For $6 a month, someone else carries that risk and I get my time back.

Melissa makes the same point from the other side: the hard part of software was never building it, but building it for public use, at scale, so it doesn't break. That's where SaaS still wins. The things that are hard to build privately are the things worth paying for:

• Reliability and scale: it keeps working, whatever changes around it.

• Built-in expertise: the know-how in the product that a home-made version doesn't have.

• Accountability for outcomes: pricing tied to results, not just access.

• Time to value: Sierra says it goes from first meeting to live in 27 days.

• Continuous improvement: the product gets better every month, while a Weekend Build stays the way it was built.

And remember the IKEA effect. If a customer has built their own version, "ours is better" won't win them over. Respect what they built, and sell to the problem it doesn't solve.

Try this with your team

1. Add the two rows. Put "The Agent" and "The Weekend Build" on your competitive analysis, next to your usual vendors.

2. Fill them in honestly. For each, write what the buyer gains, what they risk, and what it really costs them, including the time and the maintenance nobody counts.

3. Write your 25 words. Answer the question "why not just build it ourselves?" in 25 words or fewer.

4. Compare answers. Ask Sales, Product and Marketing to write their own answer separately. If you get three different answers, your buyers are getting three different stories.

Exciting, terrifying, and worth switching off for

Melissa describes this moment as both the most exciting and the most terrifying time to build software. I feel both, just not in the way you might expect.

The exciting part is personal. AI lets me build and design things myself. I can bootstrap a consultancy without a long list of contractors and still look like a professional brand.

What terrifies me isn't being replaced. My work is about people: getting Product, Sales and Marketing to tell one story about their product. AI can help with that, but it can't sit in the room for those conversations. What scares me is the speed. Keeping up with every new tool, remembering what changed last week, and still being able to switch off for my family, a walk in nature and a good meal. Those are the things that make life worth living.

I suspect your buyers feel the same. They don't want another tool to keep up with. They want a problem taken off their plate. If your positioning makes that clear, The Agent and The Weekend Build become alternatives you can beat, not threats you never saw coming.

Not sure whether your teams give the same answer to "why not just build it?" Take the SaaS Alignment Scorecard and find out where your story drifts.

Sources

• Jason Lemkin, SaaStr, "SaaS Isn't Dead. Old SaaS Is Dead. Sierra Adds $150M ARR in One Quarter While ServiceNow Drops 27% in a Day" (31 Oct 2026)

• Melissa Kwan, "This month's good stuff", your founder next door (31 Oct 2026)

• Gorilla experiment: Simons, D. J., & Chabris, C. F. (1999), "Gorillas in our midst", Perception, 28(9)