Bad Decisions on Easy Data

Success-or-Failure

Feature Ben HookwayLeave a Comment

FacebookTwitterGoogle+LinkedIn

Trying to understanding customers is an aspect of marketing that is as old as the discipline itself and all kinds of tools and techniques have been used over the years to get inside the mind of the customer.

Of course, recently, social media has given what has appeared to be a direct line to the thoughts and desires of customers. “Look at their Facebook and Twitter ‘likes and ‘shares’ in order to get an insight into their true selves” is a common technique in marketing used by many companies.

However, this suffers from a critical problem: It is straightforward to do, and so it is therefore used to inform decisions. One of the biggest dangers in any analysis program is basing decisions on data that is easy to obtain. Because easy to obtain data is abundant, organisations are lured into skewing their decision making process towards it.

Look at all the data, it’s big, therefore it must be really insightful! This is a dangerous temptation for organisations.

We had an interesting case recently with a Relative Insight client. They had undertaken a multi-channel ad campaign to raise customer awareness and action around a specific topic. Looking at Facebook comments on the brand’s page, the average number of shares and likes after the campaign was approximately double the prior 12 month average. Easy statistic to report, good for the brand at face value, so therefore the campaign must have worked?

We weren’t so sure so we performed a deep analysis of the language of Facebook comments on the brand page before and after the ad campaign.

What we found was that the language on posts had indeed changed and the topic addressed by the ad campaign was indeed being addressed in the customer discussion. However, the language was substantially and profoundly different. (At this point I should point out that we don’t do a bland ‘sentiment’ analysis of positive or negative. That is also a misleading waste of time but probably one for another blog post….).

What we found was that the discussions post-ad were far more complex and nuanced. There were a lot more emotions raised, and there was a substantial amount of discussion that detracted from the brand. These detraction discussions were shared and liked far more than the supportive comments.

The net result was that while the campaign did result in more shares and likes, the brand was damaged.

Of course, by undertaking this analysis with the client we enabled them to address customer concerns with specific content and information. This gained them a lot more customer credit and engagement.

By properly understanding customer concerns they managed to turn a potentially damaging incident into a positive outcome.

So go beyond what is easy to measure – measuring the wrong thing really well is rarely a way to get ahead.

FacebookTwitterGoogle+LinkedIn
Ben Hookway

YOU MIGHT ALSO BE INTERESTED IN