Email, web, and social media data is a great way to objectively view results. However, careful analysis is often necessary to accurately gauge what the data means. So where does that leave you when the data doesn’t make sense? If you find that your metrics defy common sense check to make sure that your analytics program is accurately gathering data.
We were recently working with a company who said that the effectiveness of their internet marketing had really tailed off. While leads and responses had remained consistent, the web reports had seen a noticeable dip that had persisted for months. With such a decline we asked what had changed when the drop off occurred but the answer was that their campaigns had run as usual. It was also strange that conversions had not dipped with traffic. Essentially the data was saying that traffic had dropped by about 40% while lead conversion had increased about 40%, a little too convenient not to raise suspicion. All the project manager could say for sure was that “We have set goals for the amount of traffic on the site and we’re noticeably off the mark, something’s not right and we need to fix it.”
The decline had happened months before our initial conversation so we reviewed the data and found that not only was there a drop off in web traffic but the hits that did trickle in were almost exclusively on weekends. As this client offered B to B consulting services, weekends were typically a lightly trafficked time.
The data flew in the face of common sense so rather than start making marketing initiatives to increase traffic; we looked a bit closer at the data itself. It turned out that a server move had disrupted their Web Trends data and a faulty setup was missing a common source of web traffic. So the project became a technical exercise in making sure that the data was being accumulated accurately.
This is an example of a larger and thankfully more obvious problem. While “inaccurate data” is often an excuse for poor results, it’s good to place a critical eye on your metrics intermittently. If you’re seeing blatant inaccuracies in what common sense would suggest, then do a technical review on your analytics to ensure they are accurately being populated. There are few things as damaging as making decisions based on false data.