profile

The JB Sales Learning Lab Newsletter

Build or buy


​.​

November 21, 2026 | Read online​

A few weeks ago, I built a piece of software in an afternoon.

I'm not a developer. I wrote a brief the same way I'd prep for a discovery call, handed it to Lovable (#sponsor), and a few minutes later, I had a working Cost of Inaction Calculator (keep reading if you want access to it), something I've wanted for years and never had a clean way to make.

Then I copied it and turned it into a deal room for one specific buyer, with their numbers and a shared action plan built in.

I was pretty proud of myself. Then it hit me that if some idiot like me can do that in an afternoon, ANYBODY can, including every client I've ever sold to.

THE QUESTION IN EVERY CONVERSATION

Build versus buy has been around forever, but it used to live with the CTO and a team of engineers. Now it comes up in almost every conversation I have, with the clients I'm selling to, with the partners I work with, and in my own head about my own business.

I've been putting content online for 20 years. You can legitimately go into ChatGPT or Claude and ask it to coach you like John Barrows, and it will. I totally understand the temptation not to pay me to come and train your team, but obviously there's a lot more to it than that.

The data says the same thing. McKinsey's latest State of AI survey found that 32% of organizations decided against buying at least one piece of software because they could build it themselves with AI coding tools, and among the companies getting the most out of AI it was close to half.

I'm not going to lie, that makes me nervous. I sell something people can try to build on their own, and I wrote a few weeks ago that I think IP is close to dead.

THE WEEK EVERYTHING BROKE

Last week, I wrote about Microsoft going down for a day. It turns out that was the opening act.

Over the next five days, the FAA halted departures across the country, Workday went offline for most of a day, AWS had a regional outage, Chase had a bug that slowed transactions at other banks, and United had to hold its own flights.

Conspiracy theories aside, my first reaction was to blame AI. Everybody is shipping faster than they ever have, people are building their own tools and walking away from vendors who have been doing this for twenty years, and things are breaking. It made for a nice clean story.

So I went looking for the proof, and I couldn't find it. Not one of those outages has been tied to AI or AI-written code, and the FAA ruled it out directly (again, conspiracy theories aside).

What almost all of them did have in common was a change. A network upgrade at Microsoft, a maintenance check at Workday, a contractor's router at the FAA. Somebody changed something in a system that a lot of other things were sitting on top of.

There is one real example of the AI version, and it's Amazon. The Financial Times reported that AWS had a 13-hour disruption last December after engineers let an AI coding tool make changes on its own (Amazon calls it user error). In March, after a run of outages on the retail site, Amazon pulled its engineers into a meeting and now requires a senior engineer to sign off on AI-assisted changes from junior and mid-level people.

Amazon kept the tools. What they added back was experienced people standing in front of anything that can take the whole site down.

WHAT I'D BUILD AND WHAT I'D BUY

That's where I've landed, at least for now. Before I decide to build or buy anything, I ask what happens when it breaks.

If my calculator has a bug, a rep gets a weird number and sends me a snarky email. If the system that runs your payroll, your email, or your flights has a bug, you're on the news.

So I'd build the things that sit close to your customer and are specific to you, because no vendor is ever going to make them for your exact deal, like a calculator with your buyer's numbers in it or a deal room for one account. That work used to sit in a RevOps queue, and now the rep who owns the deal can make it the day before the meeting.

I'd buy the things everything else sits on top of, from people who have been breaking it and fixing it for years. MIT's research found that companies buying AI tools from specialized vendors got them into production about 67% of the time, and internal builds made it about a third as often.

Experience is the one thing I haven't figured out how to vibe code. Anybody can build a training deck in an afternoon now, and I've spent 25+ years in the room watching what happens when reps try to use one. (That's the same replace, rebuild, or retrain question I walk sales leaders through, and the framework is here.)

THE OPTION EVERYBODY IS PICKING

What worries me more than either choice is the third one.

Most of the teams I talk to haven't picked build or buy. They're waiting to see how it all shakes out, and waiting feels free because nobody sends you an invoice for it.
​

That's exactly why I built this COI calculator. You put your own numbers in and it shows you what staying where you are costs every week, month, and year. It starts with the conservative number, shows the math behind every line, and tells you how much of the total came from you versus a benchmark.

It never tells you what my training (or anybody's product) would save you. I've never trusted an ROI calculator, and I wasn't going to build one.

If you want to see what it looks like inside a live deal, click around the deal room I built (fictional company, real math). And if you want to build your own version, Feel free to check out Lovable and see what you can do.

Try it on the decision you've been putting off, whether that's a tool, a hire, or what to do about your team in January.

Hit reply and tell me the last thing your company decided to build instead of buy, and whether it's still running. I'm collecting the good and the ugly.

#MakeItHappen

Got value from this? Here’s how to say thanks.

I put a lot into this every week, and I’ll keep it free. If it’s helped you, here’s how you can help me keep growing it. Pick one.

  1. Forward this to one sales leader who’d get value from it. If someone forwarded this to you, subscribe here so you get it every week.
  2. Give back. Donate $10-$500 to the Future Founder Project and help a young entrepreneur get started.
  3. Refer me. SKO season is coming. If your team needs a kickoff speaker or training, reply to this email and let’s talk.

Thank you. It means more than you know.

Update your email preferences or unsubscribe here​

© 2026 The JB Sales Learning Lab Newsletter
​
361 Newbury St, 5th Floor, Boston, MA 02115

The JB Sales Learning Lab Newsletter

Free, actionable sales advice sent to your inbox every weekend.

Share this page