Saturday, February 21, 2015

How to fuck up analytics process and benefits you could get

This is a cross-post with Super blog. Here's Finnish version of this post: Monta mahdollisuutta möhliä analytiikkaprosessi ja siitä saatavat hyödyt.

Analyst is the best friend of any executive, marketing manager, designer or developer. First change to fuck up analytics process and collecting useful data, is the planning phase. If you ignore analyst in this phase, you will ensure that measuring is far more difficult in the next phases. On the other hand, if you don't have any goals, then it doesn't matter. If you don't have goals in your business, you can stop reading right now! If you do, please continue.

Keep your analyst aka right arm with you

When you are planning your marketing campaign, KPIs, raising the budget, newsletter, redesign of the website or something else, remember to take your right arm (analyst) with you (or the vendor which is responsible for measuring and analyzing). I have seen lots of failures because the person responsible for measuring and data collecting gets the information of plans too late (or never). The sooner, the better. In this phase it's ideal if you are able to tell the goals of the plans. These goals are not impressions, clicks or visits - they should be sales goals. If you think that digital marketing and visibility don't have anything to do with your sales - you can stop reading.

Don't forget measuring tags in technical implementation

With your analyst, it should be quite easy to draw on the white board the optimal user experience. After this it's quite easy to plan the measuring of the micro and macro goals. This way you are able to get the needed analytics tags in the implementation phase among other code production. If you ignore the analyst in the planning phase, it's possible that measuring tags are forgotten or analyst is creating some tags when there's already a rush to publish the campaign / site / newsletter / what ever. If you are using for example just the basic tracking script, you are not collecting the right and useful data. If an analyst has been in the planning phase, she can talk with developers about measuring and the different ways to implement the tags. This is a way to make sure that the implementation of measuring is done right.

Forget testing, forget quality data

Before launching things, you_should_do_some_testing. During this phase you should test that the data collecting is done right (according to measuring definitions done in the planning phase). If you are not testing at all, you are not able to trust on the collected data 100 %. It is not rare that testing phase is skipped. If you are not auditing your data, you don't know if you are making important decision based on bad data. During the testing phase you could also create needed reports. But beware: in a report, there may be just numbers - and you are missing an analysis.

There are so much useless number reporting

You are not able to get useful information without analyzing data. When campaign has been launched or any change has been made, you are getting data in (if you haven't forgotten measuring). If you follow the process, you have a change to save new quality data. Next, you have to analyze data. Analyzing should be in line with heart beat of the business. The more you have traffic, the faster you have a chance to analyze for example ads, landing pages and (purchasing)funnels. So reports may include just numbers but analysis has also text (hopefully telling you what the numbers mean). Analysis should be done against goals, your previous performance or compared to competitors. Output of the analyzing should include development, change or testing ideas!

Decisions, decisions

One of the most important task executives have, is to decide. Output of the analyzing phase is the input for decision making. Decisions are the input for redesigning. Decision can be big or small but it's very important that you are making those decisions. When you are able to manage this analytics process, you are also able to achieve constant development cycle and data driven culture. The more you have traffic and actions (heart beat of the business), the more faster analytics process should go around. The more faster the process is rolling and the more faster you are able to implement development ideas, the more faster organization is getting useful information and experience. If you haven't done, for example within last year, any A/B or MVT test, it's fair to say that your analytics process has failed!


We are having a Super seminar with Ensighten in Helsinki on March 13. If you are around and would like to see either Ensighten in action or me telling more about analytics process and how to measure the process, sign up here.