Skip to content

CRM Guide - Evaluation

1. Types of system

 In my opinion there are four types of systems:

  • Custom build - this is a totally bespoke system built just for you, typically in partnership with a development agency
  • Platform build eg Salesforce, Dynamics, CiviCrm - while these can just be configured, they are complex and would always require a partner to work alongside you. They also offer custom coding elements inside the platform.
  • Low/no code eg Airtable - depending on your in-house skill level, you might be able to build this out yourself. These don't come with built-in datasets, ie. there isn't a People, Company, Case container, but often there are templates you can use and the interface is straightforward.
  • Configure off the shelf eg Beacon, Donorfy - These systems are tailormade for housing all the data you need, and come pre-configured to cover the bulk of scenarios. You then configure on top of this to get it just right for you. You have the option to DIY or work with a partner.

 Should we buy or build - well we've written a whole article on that one! 

 We've also built out a long list of charity databases here. 

2.  Reduce your longlist → shortlist

 You only have to Google (or ask AI) for a list of charity CRMs and you will probably have a list of six to ten systems. It's now time to whittle that down - again AI might be a useful tool here. Feed it your longlist, user personas, user stories and brief, and ask it to provide a shortlist of three. Be sure to ask it to provide reasonings, not just for the recommendation but also why it has discounted some systems.

As always, AI is a weapon in your armoury here and you should also conduct your own assessment.You are essentially trying to shortlist based on which system fulfils most of your user stories.

Once you have your shortlist, send the companies your brief, personas, user stories and ask them to respond and provide a demo. 

3. Preparing for Demo's

 Don't make the mistake of thinking you can just rock up to a demo. You need to put a bit of ground work in, before and after the demo, to make sure you:

  • Have the right representation
  • Have prepared the provider in advance for anything specific you want to see
  • Come prepared with your questions
  • Conduct a centralised debrief

 We have written a separate guide on attending demos. 

4. Making the most of trials

 Most providers will give you a 14 or 30-day trial of their software to allow you to evaluate it - it's important to make the most of that trial. You should approach the trial with the same rigour as the rest of the stages here. Having a play about is important, but it's not enough. Here are a few things to think about:

  • Get the timing right - there's no point having a trial if no one is available to look at it
  • Create a plan to be clear on the who, when, what. Don't make this one person's job; you want feedback from a few people
  • Go back to the feedback from the demos and the responses from the provider to your brief. Are there certain aspects that this system doesn't do as well, or where they have provided a work around? Spend some time reviewing that information and then make sure you look at this during the trial. Can you compromise here, or will this be a deal breaker?
  • Don't burn time on the basics like 'does it hold people records?' (they all will) - what matters more is what that record looks like, how easily you can see all the information connected to that person, how you can see the history of interactions with that person.
  • How are you going to collate feedback from the trials - what is your scoring criteria that will feed into how you choose your final solution?

 We've provided an example scoring matrix. 

5. Making your decision

 The scary bit right? Yes and no. You've done all the hard work - now you just need to trust the process.I would recommend two lots of scoring:Against the user stories - for each provider add a score to each user story where:

  • 0 = can't be done

  • 1 = can only be done with a workaround or other integration

  • 2 = can be done

Then add up the total score and divide that by the total number of user stories x 2 (max score) x 10 to get you a total score out of 10 (this sounds worse than it is!)

For example, you have 100 user stories, so the maximum score would be 100 x 2 = 200.

System 1 scored a total of 150

(150/200) x 10 = 7.5

Meaning system 1 scored 7.5 out of 10 on the user stories.

I would probably only include the Must and Should user stories in this calculation.I would then score as part of the trial process (using the example scoring matrix). The score from the step above would be the score you put in the functional fit, then you complete the others to give you an overall score.

The score is just part of the decision though. It's important but it's not everything. Let's say there were just two points between two of the systems, but one of those systems is missing a mandatory requirement whereas the other is missing a should requirement - this might sway your decision. Or maybe the overall scores are the same, but one scored much lower on security and much higher on usability. Again this might play into your decision.