Integrate your own custom suite to suit your processes, maybe, but never build from scratch. (And we should not have to be talking about this again after just publishing on the subject two weeks ago, but too many conversations are indicating that we still need to shout this loud and clear!)

For some reason, this comes up every decade, usually after a hype cycle has peaked, marketers have switched from focussing on solutions to sound bites from a suite of providers who have released products that don’t meet customer needs, the implementation failure rate has edged back up to the 80%+ range, and customers have gotten absolutely positively fed up with the whole situation.

Customers, fed up with the valueless hype, marketing sound bites, high failure rate, and utter lack of solutions from the vendors targeting them on a daily basis, start to think that the right solution is to build their own.

Sourcing Innovation tackled this subject in depth back in 2015 when it wrote a 4-part series on why you should NOT build your own e-Sourcing solution, followed by an explanation of why you should not build your own Contract Management and e-Procurement platform. (links here)

That’s why we are both repeating and elaborating on last Friday’s Rant on why A Company Should Never Build It’s Own Enterprise Software Systems.

Not only do we have the situation where:

• the company is not an expert in building software products
• the company is not an expert in best practices across all of its processes
• by the time a custom solution is developed, it’s out of date
• it’s not about the product, it’s about the process you should be working toward and, most importantly,
• it’s about the data that drives the process!

But we have the situation where, as highlighted in THE REVELATOR‘s article:

1. Developing your own is NOT being an early adopter! (Which is what many companies considering build-your-own think they are.)

Early adopter means someone who adopts leading edge technology from a third party, not someone trying to fast track their digitization effort with custom built tech. This is just high risk with little chance of reward for all the reasons mentioned in all of our prior articles.

2. They think Gen-AI will fix their data problem and allow them to develop their own!

If you’re read anything on Gen-AI on this blog you know that’s the last thing it will do. For Gen-AI to have any chance of working at all, it needs a huge amount of good, clean, data. Otherwise, it’s garbage in, hazardous waste out. No technology has ever needed such large amounts of near-perfect data to have even an abysmal chance of working, and the fact that the marketing madness has convince many CPOs that Gen-AI can fix a data problem is downright terrifying!

3. They obviously think that the initial quote will be close to the final cost.

No where are cost overruns more extreme than in custom development by a non-software organization that contracts a Big X with poor specifications that look easy, and that, due to lack of manpower, sends The C-Team (if you are lucky) because it’s just another instance of system X (when it’s not).

To be honest, in this situation, if the costs ends up being only 3X to get something usable (but still not what you wanted), given the high technology failure rates, that would be amazing.

We know it’s hard to find appropriate solutions given all the noise out there, and the overabundance of vendors that all look, sound, and go all in on useless Gen-AI the same, as it just takes one glance at the Mega Map to figure that out, but that doesn’t mean there aren’t vendors out there appropriate for you. Vendors that put solutions, not tech first, that built affordable tech that works (and didn’t take too much money from investors who then insisted on quadrupling the price), and that will work in an ecosystem with out vendors to solve your problems.

You just have to look hard. Real hard. Probably harder than you’ve ever had to look before. (Expect to eliminate 6 out of every vendors you look at for short list consideration and probably go through 20 to find 3.) But trust us, when you find the right vendor, it will be worth it. The solution will work, will configure to your liking, will be extremely usable for the problems your team faces every day, and will be one where the provider will grow with you for the decade to come.

Good things come to those who wait to find the right vendor. (Even if they have to crawl through multiple pig sties to do so.)