HMRC and an example of the problem with defaults

Since February, so for about five months, I’ve been working on connecting to HMRC systems to handle VAT.

It turns out that this is not really what we’re supposed to do.

Yes, anyone can apply for credentials, but it’s built around two scenarios only: big companies, and software developers who will build and sell a tool. The hurdles are put in place with that in mind, and something internal for a handful of people … doesn’t really compute.

So I do what they say, wait a couple of weeks, then get a response telling me that they need something else that doesn’t make much sense in our context, to which I respond typically the same day, then wait again. I like to think of it as demonstrating persistence.

The point this makes to me is that the modern world is built this way, around bigness, gatekeepers, and defaults.

When government does something, it assumes the actual handling of it will be through accredited intermediaries. When it talks to business, it talks to big business. When it opens up, it assumes that those who engage will be of the kind they assume qualified to do so. Why would anyone whose business is not in some way “dealing with the government” deal with the government?

It’s easier if everybody sticks to the big obvious options, not just for the government.

If you want to search the web, almost everybody uses the same search. Except us, because we pay for Kagi.

If you’re a small business, your accountant will probably suggest the same software they suggested to me. And that I’m avoiding by stubbornly building our own.

If you’re exchanging documents, they’re probably in a few proprietary formats.

If you’re looking at ERP software, regardless of your actual needs, you’ll feel the pull of the big names, the obvious choices at your level.

If you’re implementing an ERP, you’ll be steered towards best practices, built-in modules and approved third parties.

This all makes perfect sense, from the most obvious perspective. If something is best, then let everybody use it, and everything is simple and efficient. Everything is compatible, everything is known, everything is as good as it can be because the best was chosen and nobody is missing out.

So, why not just do that?

If there’s accounting software everybody likes, and it can handle talking to HMRC, why make life difficult by doing something else, let alone doing it all manually and bespoke?

Well, for one thing, there’s a principle at stake. When one “best” is allowed to become not just best but only, “best” becomes a slippery thing. Something can technically only be best if there are alternatives, and when there aren’t it tends not to even stay good for long. We see a lot of this around. Entire business models are based on becoming the choice, and then squeezing that market.

So as a personal thing, where I can, I feel obliged to keep alternatives going. Like a minority search engine. And I pay a price for that, but also they have to work harder to be worth it for me.

But really, isn't "best" a slippery thing even in ideal circumstances?

There may be particular reasons why what's best for you is less than ideal for me. There may also be reasons why I accept less than the best for the sake of ease and compatibility, but not always. Sometimes I really do need the best, and the best is not what is easily available.

That, honestly, is not the case with accounting software or connections to HMRC for us. But after the last solution I chose simply closed one day, the perceived cost of being locked into someone else's way of doing things rose enough to make flexibility more important. It's not so much that we needed to do it from scratch for the result, but that we needed to know what we could only find out by doing it from scratch. We needed the knowledge that we can be independent when we need to. We can do things for ourselves. So when we do need the best, we know we have the choice.

Choice is good, choice is needed

A lot of things we've done for customers have been because they needed another choice.

The standard Epicor web-commerce connector wouldn't work for them, so we worked with the web agency to build what they did need.

The product configurator module was overkill, so we created a light-touch alternative.

In each case, we were careful to explore the standard routes with the company, presenting the case for "best practice" and the usual methods. But when it proves to be justified, you do need another way, whatever the pressure to simply do it as normal.

And when that's the case, what you need is the stubborn sort of company who's done things the hard way to prove it can be done. Because even that sort of company is a needed option among the many who do things the "best" way.