Things we learnt trying, and failing, to build a ‘CRM for everybody else’

A fantastic guest post by choosenick from Makeshift we’ve been allowed to share.


A year ago we set up Makeshift, a company dedicated to ‘giving a leg up to the little guy.’ Over the past year we’ve invented and built 11 products which have met with varying success. Over the next few weeks I’m going to try and write about the some of major things we’ve learnt in this first year of operation by telling the story of these products.

One of the ideas Paul and I had been bouncing around before starting Makeshift was ‘something to do with lists of people’. We’d both realised that we both maintained lots of lists of people for varying reasons, and we’d both considered using CRM system to manage these lists but found them far too focused on sales and closing.

In Paul’s case he used Twitter lists to keep his lists in shape, and I used a mixture of spreadsheets and desktop notes. I was mainly making lists of people to work with or hire, and he was mainly making lists of people to meet up with in and around the startup scene.

Independently we’d both come to similar conclusions:

  • Making lists of people is a common task
  • Lists of people are different to lists of other things because the people do all sorts of stuff online that you want to keep track of.
  • Current ‘people listing’ apps were flawed for everyday use because they either forced you to act like everyone was a customer (CRM) or they wanted you to convert the people you add into users of their websites (social networks).


Above: a bit of our ‘business case’ for Listerly

CRM for normal people

It felt like there was a really open goal here (and to be fair I still think there is) to create an app that was like a CRM, but for normal people. Independent of any one social network, but connected into the social web.

“An address book on steroids.”

Could we make an app that was just a great way to make a list of people?

Maybe, but as we discovered with Bitsy, the problem was that we didn’t optimise for a specific enough use case – for example tracking freelancers or building marketing targets – and instead tried to go big before we even started.

So instead of focusing we went broad – but this design results in a massive app (even if the desired UX is very small and tidy). You have a really complex data model, with an infinite number of identities associated with individual people, who need to be able to claim their ‘person-ness’ and then authorise their various identities (e.g Twitter, email, phone).

Other people have tried variations on this, and some are doing really well. LinkedIn tries to gobble every identity into a CV. Rapportive scrapes a lot of stuff to try and make your contacts book smart. Full Contact tries to offer identity as a service. All these are really ambitious tech plays – whereas we had basically zero resource.

Er, who’s working on this?
We started working on Listerly in week two of Makeshift, at the same time as revving up on Bitsy, and getting through all the hassle of setting up a new company. This meant that, essentially, we didn’t have anyone to work on it, so it sat as wireframes on the wall for a bit.

Above: Some UX type stuff

Fortunately, Jon joined us after a few weeks, and we set him to work on this as his first hack. Oh, and he had to help with Bitsy. And doing the Makeshift site. Hmm. He made some good progress, and made the (at the time helpful) suggestion that ‘maybe we’d like to make this using this new Meteor thing’ that he’d been playing with.

Technology selection fail

At the time we had no technology stack strategy, so we just said, alright, give it a go. We knew we wanted to explore a javascript development framework, but I think Stef and I missed the fact that Meteor was a fully hosted service at the time. Anyway, Jon started hacking away, and suggested that we ask his mate Chris to come and help out too.


Enter Chris
So Chris came on board, but only part time, and we asked him to basically do everything – product definition, UX planning, UI design and full implementation, on a brand new technology platform with no documentation. On three days a week. With no help from our senior team.

Unsurprisingly, despite his best efforts, Chris struggled with this unfair brief. Along the way I was extra unhelpful piling more and more feature ideas on him – and Stef was unable to help because he didn’t know what the hell Meteor was.

Exit Listerly
The end result was the Chris and Jon did, heroically, ship a Meteor app that did something useful – let you build really basic lists of people using mainly twitter as the identity hook. When they demoed it at the Meteor meetup group there was gasps – ‘you guys have built something useful on Meteor! No way!’ (I’m paraphrasing).

But, as you can imagine, the technology choice, the huge ambition, the lack of use case definition, plus the massive under-resourcing had doomed this project and it never really made it ‘out the door’ properly. Me, Stef, Jon and Nathan were all concentrating on Bitsy and Paul was scheming up the next product, Help me Write.

“The massive under-resourcing had doomed this project and it never really made it ‘out the door’ properly.”

What did we learn that we took forward?
Well, we didn’t learn much about the market or customers here because we never really even got our MVP out the door to test a hypothesis properly. Through the process of designing it, we did figure out that the use case of ‘lists of people’ is probably too big, and that it needs to be ‘lists of freelancers’ or ‘people I want to call on my phone’ or something.

We did learn a lot about how to ensure projects get the right amount of attention, and that we should make technology stack selection decisions more carefully.

What’s next?
Listerly was an inconclusive fail because we never really shipped it, and the reasons were purely down to internal execution failures, not any really conclusive market insights. Damn. We spent around £7,000 on it all in, stretched out over six weeks or so.

Personally I would love to go back to Listerly at some point. I think there’s a thing there. I think this because my desktop currently looks like this:

If you feel my pain and would like to use a ‘CRM for normal’ people I’d love to hear from you – email me nick at makeshift dot io – if we get enough requests we could always revive it!