Centralise at your peril

With thanks to Steve Messer, a brilliant product manager. This post is better with a basic understanding of Wardley Mapping.

I’m going to go further than Steve and say that if you’re anyone, anywhere, whose service is based on centralising anything: be on the lookout for the next best thing. Be aware that you may not be able to survive it.

The urge to centralise any function in your organisation is probably coming from the fact that you’ve noticed you’re spending an absolute fortune on everyone doing their own thing. Maybe you’ve finally broken down the ‘documentation’ line item on your budget and discovered that every team in your organisation is hiring their own typists, training them themselves, and buying their own typewriters. No two typists are the same, and no two typewriters are the same. Also, apparently someone’s still writing things on – goatskin? Is that right? Is that allowed? Yikes.

A Wardley Map. The top node is "A nicely typed document", and it's towards the genesis side of custom. It's supported by "effective typists", which is towards the product side of custom. This node is supported by "typewriters", also custom, "training", which is more custom, and "nice young ladies of an appropriate age and temperament", which is at the edge of product/commodity. This node is supported by a commodity belief: "the sexist attitude that lets me get away with saying dumb shit like this"
Luckily for us, in the present day, that sexist attitude is completely gone from the map

Now, despite the fact that you and I have fallen through a wormhole to back when typists were a thing, your instinct will probably be the same as it would be today: to reach for the big hammer marked ‘centralise’. Let’s take a second to make sure we’ve got a nail in front of us, rather than a box full of elephant dung.

Centralising the ‘typing’ capability is going to massively bring down costs, true. You can also centralise typist training, and take out a single massive contract so that each typist has a typewriter. Teams will no longer have to fight finance to get their typewriter costs approved, and they’ll have things typed whenever they need.

The map above is reproduced, with the following additions: "A nicely typed document" now relies on a "typing pool", which is a product capability moving towards commodity. This is supported by "effective typists", now a product commodity, which is still supported by "typewriters" and "training" - both now products. Nice young ladies and sexist attitudes stay where they were.

Of course, now that it’s centralised, you’ll need to hire someone to manage the queue of work coming in. And the typists were mostly learning by doing, so you’ll need to hire someone to train them properly. And you’ll have to make sure nobody’s sneaking around the new centralised policy and hiring their own typist, unless they’re very important. Obviously.

Now that we’ve got all the typists together, you’ve noticed that some of them sometimes don’t have any work. In which case, perhaps they’re not needed? Efficiencies! Trebles all round!

Unfortunately now wait times are creeping up. More people want things typed, now that you’ve got this fab new central resource. Rather than hiring someone, all the new teams are funnelling work to you. People are waiting longer for work. They’re starting to get frustrated.

The map above is reproduced, except the top node "A nicely typed document" now also says "...turned round in a day or so". This reflects the increased expectations of users.

You get an expense report for a new typewriter. It’s 50% cheaper than the ones you procured centrally, even though you bought 100, even though you were assured it was the best price possible. It’s why you agreed to it for the next ten years. That was last year.

You set the expense report on fire and use it to light your pipe. Your hand shakes with rage as you do.

The next day someone asks for blue ink. Blue! You dash off a letter to The Times, decrying the youthful rush to gaudy nonsense like blue ink. You have one of the typists type it up. It takes two days. You grit your teeth so hard you hear something crack.

You have a chat with your equivalent in the firm down the road. He decries the spiralling costs of typewriters; the huge numbers he’s buying; the dizzying variety of inks – not just blue, but red as well. He had to buy a typewriter with italics the other day. Italics! You smile smugly into your gin but it’s only when you leave that you remember he didn’t answer your question about how much money his firm was making but that his shoes and hat were both new.

If you’re smart, you reverse the centralisation decision within the next six months. If not, your company goes under five years later and vanishes without a trace.

Let’s come back to the present day. What have we learned?

Centralising an expensive resource is often a good idea. You standardise inputs and outputs, and can run the resource at full capacity. However, standardisation results in rigidity. That’s not necessarily a bad thing – if you can be confident in your predictions for the next ten years or so, secure that deal. Accept the rigidity, knowing you won’t need to flex the capability.

However: centralising a rapidly-evolving capability, or one that’s subject to significant change, is a dangerous gamble. If you’re really keen to do it, my recommendation is to map out the capabilities and examine which ones are most subject to change. If you’re great at hiring typists, for example, can your process flex to hire a plumber? Or a user researcher? Or a strategic consultant?

Plan for change, and try not to be over-optimistic about how standard your processes and tools are. Good luck.

Leave a Reply

Your email address will not be published. Required fields are marked *