The Psychology of Harvesting

When I’m out talking to people who are creating models, the one thing I almost always end up talking about is “Harvesting”. This is how I refer to the deliberate pulling-in of information into a model, not as part of a specific project, nor as the basis for reverse engineering (though these are clearly useful as well) but just because it might be useful later.

So why would you spend time pulling information into your model, just for the fun of it. You’ve already got enough to do, right ?

I’m aware that I see a non-random sample of models and modellers. If you’ve been modelling for years and have large, multi-project, long-lived models which are integrated into your development process, then we’ve probably never met. You don’t need modelling help. (I’d still welcome your views).

So I see a disproportionate number of teams who are just getting started in producing models. They are typically a proof-of-concept, trail-blazing project, which they hope will lead to more and better projects in future. So they are really focused on making a success of their project. And that’s great – I’m happy to help.

Something else they are usually keen to do is to spread the word to the rest of their organisation about what the’re doing. It’s often one of the objectives of the project. And this is where the Psychology of Harvesting comes in.

A Little Story

A few months back, I visited one such team. They were a smart crowd, and they had done the training on their tools & modelling techniques, and had gone on (with only a very little help from me) to produce a really useful model. It had processes and use cases, risks and issues, domain models, and even some simple wire-frames for the application they were designing. Having it all in one model had delivered some really useful insights, and they had spotted some ‘gotchas’ which might not otherwise have been found this early in the project. They were happy, their project manager was happy, and their stakeholders were OK too. (stakeholders = just OK is a good result!).

So they went out to spread the happiness elsewhere in the organisation..

…and the results were poor.

Far from being embraced by other teams as the saviors of the company, they found indifference, confusion, and in a few cases, outright hostility to the idea of modeling.


So the team leader and I sat down to try and work out why this had happened. After all, her project was a success story, so naturally others would want to do the same. Wouldn’t they?

After much coffee, we decided it was a combination of two main factors:

  1. Fear of the new – this is a well understood, natural human reaction to change, so these guys tried to help by offering training and support for new teams. They had even established a small fund of money to pay for it. But that wasn’t enough. How to persuade them to take the training and support in the first place?
  2. Strangeness. When they showed other teams their model, they were rightly very proud of it. But the content of it was largely un-familiar to others in the company. Maybe they recognized some of the words, but it was someone else’s project after all, so all a bit strange.

We realised that the solution to overcoming the strangeness was Harvesting.

Harvesting to Persuade

So they tried a new approach. Rather then going to new teams with their own beautiful model, they took whatever they could find which related to the domain of the other teams. Maybe their lists of requirements, or their project risk register, or just a key diagram which gave an overview of the project. Or just the boilerplate ‘2 paragraphs about this project’ that they put at the start of all their documents. Or some screen-shots from an existing, related system. Whatever they could easily find and pull into the model.

In all cases, the goal was to let other teams open the model, poke about, and see something familiar.

The effect of this was immediate. Here was a strange new way of thinking and working, but with their own knowledge inside it already. Here were ideas and terms they recognized, presented in an unfamiliar style, but  definitely stuff they knew about.

And this seemed to make all the difference. They were now keen to understand where all the information came from, how the different bits related to each other, and they immediately started to think about ways to link information together to get a bigger, richer picture. They ‘got’ the idea of a model.

So this is the Psychology of Harvesting. Maybe not a super-advanced technique – probably anyone could work it out themselves given time – but a really simple and effective way of spreading the word about modelling.