Implementation Governance

by Mar 16, 2020

In the last post of this blog series on being success­ful with digital analyt­ics, I talked about how better docu­men­ta­tion can help your end users under­stand your imple­men­ta­tion. While you may take your imple­men­ta­tion for granted and know all of its ins and outs, your busi­ness users don’t have that same luxury. Just a small amount of effort towards docu­men­ta­tion can go a long way. In this post, we’ll discuss how better imple­men­ta­tion gover­nance can also help your end users.

When I say imple­men­ta­tion gover­nance, I am refer­ring to how orga­nized your imple­men­ta­tion is and how much of an effort you have made to make it easy for your users to navi­gate the analyt­ics tool. Most digital analyt­ics tools provide some admin­is­tra­tion features that can be used to main­tain the imple­men­ta­tion, but many orga­ni­za­tions fail to take advan­tage of them.

One of the biggest mistakes I see orga­ni­za­tions make in the gover­nance area is having incon­sis­tent data sets across differ­ent websites or mobile apps. It is best prac­tice to have your differ­ent analyt­ics data sets mirror each other so you can easily merge data and re-use imple­men­ta­tion compo­nents like reports, dash­boards, segments, etc. Here you can see an example of an orga­ni­za­tion using the same data slot for differ­ent things, which will likely haunt them down the road.

Dash­board orga­ni­za­tion is some­thing else you can do to help your users. Most tools allow you to create dash­boards, but I suggest that you create a dash­board of dash­boards and make that the start­ing point for end users. This dash­board of dash­boards is like a table of contents of your imple­men­ta­tion and can be config­ured to make it easy for end users to find what they are looking for. Here is an example of this concept that I saw at an Adobe Analyt­ics event:

Another thing you can do to help your users is provid­ing a way to denote which imple­men­ta­tion items are “approved” or “blessed” by the analyt­ics team. It is not uncom­mon for users to look at segments or calcu­lated metrics and see hundreds or thou­sands of them. Some­times there are even ones with the same names!

In these situ­a­tions, it is impos­si­ble for end users to discern which ones have been vetted and should be used. To remedy this, some analyt­ics tools have a concept of “approved” metrics or segments where an admin­is­tra­tor can flag them as being correct. Here is what it looks like in Adobe Analyt­ics:

If your tool doesn’t offer this feature, the other old-school option is to simply create a fake user login and use the name “Analyt­ics Team” and save all of the reports, dash­boards, calcu­lated metrics and segments that are offi­cial under that ID. If users see the owner as “Analyt­ics Team” they know that it can be trusted.

Speak­ing of dupli­cate items and having way too many metrics and segments, I recom­mend that you review all items in your imple­men­ta­tion for stuff that can be removed at least every six months. Since most analyt­ics tools do a poor job of telling you where these items are used, I wouldn’t delete them, but if you find stuff that you think should be deleted, change its name to some­thing like “TO BE DELETED ON 6/1/20 — “ in the begin­ning of its name. Then tell your users that if they ever see some­thing with that nota­tion to contact you and you will help them swap it for the offi­cial version of the item.

Another feature that has made its way to some analyt­ics tools is tags. Tags are ways to clas­sify imple­men­ta­tion items in a way that is similar to tags for blog or social media posts. For example, if you have a bunch of metrics or segments related to mobile apps, you can tag them as “mobile” and later search for them by tag. This can be very useful if you have an espe­cially large analyt­ics imple­men­ta­tion.

One clever way to use tags is to include all of the elements contained within a calcu­lated metric or segment as tags. For example, if you have a segment that includes a metric and a dimen­sion, you can tag it with their data point names. Doing this allows you to later filter items by data point name and find any items that include that data point:

These are just some of the ways that putting some fore­thought into your analyt­ics imple­men­ta­tion admin­is­tra­tion can help your end users.

Action Items

Your home­work for this post is to:

  • Create a dash­board table of contents. Don’t forget to include a link to a dash­board that contains your data dictio­nary from the previ­ous post.
  • Pick a method to commu­ni­cate which imple­men­ta­tion items are “approved” and mark all approved items for your users.
  • Go through all of your imple­men­ta­tion items and either remove old, outdated items or dupli­cate items or mark them for future dele­tion so you can work towards their removal.
  • If your analyt­ics tool has tags, try tagging as much as you can and see if that helps users find what they are looking for. If you have a large imple­men­ta­tion consider adding the data point names as tags to make it easier for users to narrow down what they are looking for.

In the next post we will talk about the imprtance of exec­u­tive support and how to attain it.

We’re here to help you through this.

I consent to having Search Discovery use the provided infor­ma­tion for direct market­ing purposes includ­ing contact by phone, email, SMS, or other elec­tronic means.