At my company, the major issues are employee retention, time wasted doing data processing, and job satisfaction. The analytics department here isn't so directly connected to any hard and simple KPIs like revenue. We're an advertising company, we report on the effectiveness of ads which has lots of variables, from the types of art in the advertisement to how it's ran. So for us, Alteryx was a solution for our internal office problems rather than a way to drive a simpler KPI.
To justify Alteryx when there is no simple KPI, you have to generate the data yourself. You know what kind of things Alteryx is helping with, so you should derive a way to track that data. For us, the solution was a pre and post Alteryx survey.
We had two major goals:
- track improved time for data processing (perceived and actual, I'll get to that later) - check to see if Alteryx makes people more satisfied with their work, which would affect retention
To measure the effect of Alteryx, we created two surveys. The surveys recorded anonymous but basic data about the user and asked a bunch of questions to track above. The questions rated on an ordinal scale, so we could easily compare before and after responses.
Before anybody was allowed to have a license from us, I required them to fill out this pre-survey without telling them there was a post-survey. Once they completed the surveys I would give them Alteryx access. Three months later (or whenever I knew they were more mature with Alteryx) I would reach back out and have them complete a post-survey.
Some of the questions:
- How many hours per week do you spend on data processing? - How many hours per week do you spend on pulling data? - How many hours per week do you spend searching for data? - How satisfied are you with the workflow processes you do regularly? - How satisfied are you with your job?
By asking them the same questions before and after, you can measure actual change. There's of course the issue of the users reporting accurate numbers, since you can't put stuff on their machine to track if they actually spend x number of hours on processing, cleaning, whatever. Since the users are self-reporting numbers, the data can be influenced by their opinion of how much time they spend before and after Alteryx, but in my opinion, this subjective response is just as legitimate. It shows how the users feel Alteryx is actually working, even if they get the numbers wrong in both surveys.
If you record this data, you'll be able to go into presentations with executive leadership with more than just hearsay. Doing a survey and collecting other data affected by Alteryx is the key to making Alteryx the central tool in your work processes.