Our Data Sucks. What Do We Do?

This blog is a part of a greater series where our readers are submitting their questions so that we can answer them right here on the blog. Don't worry, we are not sharing the name or organization who submitted the question so that everything is anonymous and generalized, but we know that if they have this question, other readers do as well, and we want to deliver the best content right to your email box every single week. 

This week's question says, “Our data sucks. Most members have not filled in their profiles, and we are not sure where to begin as we look to clean up data and make better decisions in the future. What should we do?” 

First, I want to normalize that this is extremely common. 

Out of the dozens of organizations that we work with every year, almost every single organization makes this statement. And the few who feel good about their data admit that their data is lacking and not where they want it to be. 

Let's all exhale together, and recognize that as an industry, we're in the same boat.

Yes, we want to move forward, but we're not behind our peers. 

Next, I want to encourage you that it's better to have correct, but minimal, data than to have tons of incorrect data.

Imagine if you had fields completed that were outdated by 10 or 20 years. You’d have to go through and scrub the data clean, or get rid of old entries so that you could have proper segments.

This should open your eyes to the fact that data is an ongoing initiative for every organization, and it is not a one-time job. 

Now, if you are lacking data, here are my best tips of what to work on so that you can have clarification around what value you should be offering and, from there, be able to determine what price you might be able to charge for that value. 

1. First, you will never have 100% of your database give you 100% of the data that you're looking for. 

   Your goal is to get a good sample size. What this means is you need to limit the scope of how much data you're asking from the people in your database. Yes, it looks great to have 20 different fields, but the reality is most people are not going to fill it in, let alone update it annually, otherwise they would have done it for you already.

   My first recommendation is that you narrow down what segmented questions are most important to ask your audience and focus on those.

   This typically will be three to five questions, at most, that will help to create a persona around that individual. You can meet as a team internally to decide what matters most.

   Typically, these can be things like their ‘level of seniority’, ‘if they have a specific certification’, ‘how much education they have’, ‘how many years of experience they have in the field’, ‘If they are a purchasing decision-maker’. 

   Once you determine which few points are most important, I want you to focus on just getting information there. 

2. Second, you need to make this really quick and simple for your audience. 

   That might mean that you might have to do more manual work on the back end. Yes, you can send them to their profile to update it, but unless your going to strip away all of the other questions on their user profile in your AMS, you're still asking them to complete the same 20 questions. 

   What this means is you might need to send out a survey with just the three to five questions and ask them to enter the professional email that they use as a member, and then complete those three to five questions, and then your team goes back and does a CSV upload into their profiles. 

   If your AMS doesn't have this feature, it might mean that someone has to manually add it, which will take time, but again -- if you want great data, then that's something that you have to consider. 

3. Third, consider what sample size is enough of a response to get an idea of who is active within your database?

   Notice that we use the word ‘active’. 

   Again we're not looking for everyone in your database to fill in their member profile because that's not realistic.

   Look at your click rates on your emails. Typically they're going to be 1 to 3%, maybe higher depending on the content. I would shoot for a similar size of your audience to actually click through to potentially complete these three to five questions.

   So if you have a database of 30,000 people, then you're probably looking at 300 to 900 people clicking through to potentially give you this information. If you get a higher number that's great.

   Don't forget that when we're comparing completion rates to your click-through rate, that's the number of people who click. That doesn't mean that they complete the survey.

   So if you have 30,000 people and 3% click on it, you would have 900 people click through, but then the number of people who actually complete the survey might only be 30-50%, so you still might only get 300 to 450 people giving you answers for these personas.

4. A bonus tip: include these three to five pertinent questions, as appropriate, on registration for events or for digital education opportunities. Again, have an intentional effort to then move that CSV export over to your AMS to update profiles. 

Is there an easy answer to this question? No.

If I had an easy answer, I would be a millionaire for solving one of the biggest problems that associations face. 

But is there a pathway forward that will take some elbow grease and intention, but get the results that you need so that you can get a better idea of who is in your audience and how you can best speak to them or create personas based on who they are?

Absolutely.