Let’s say you’ve walked into an existing situation, whether it’s a new job at a company, a promotion to project leader, a new client you just took on. What’s the first thing (after understanding the goals and objectives) you should do?
Look at the measurements for the trends.
Start by grabbing the top 2 or 3 KPIs and getting up to 12 months’ of historical data on them. Grab them, put them in your favorite spreadsheet, and start looking for trends.
Here’s a quick way to do this. We’ll use my web analytics as an example, but you can apply this to any KPI that is in good condition (data collected correctly and without errors). First, decide what constitutes a reasonable moving average for your data. For example, when I’m looking at website traffic, I like to measure by the week for business blogs (that smooths out the rapid drops in weekend traffic). In turn, because businesses run most often in quarterly cycles, I think a 13 week moving average makes a great deal of sense.
If my blog was consumer-facing, I might choose a different moving average that more closely fits how consumers behave. The average business cycle for a consumer is typically more monthly (rent/mortgage, car payments, etc. typically occur monthly) and therefore a 4 week moving average might make more sense.
So here’s my blog’s weekly and 13 week moving averages:
Now if your number crunching software allows it, add in up-down bars:
Look carefully. There are a great deal many more misses (weeks below trend line), especially in the second quarter, than there were hits (weeks above trend line). If this were a full-time business, then there would be heads rolling in the cubicle farm.
Once you’ve gone through the 2 or 3 top KPIs, find the one trend that is in the worst shape. This is a really critical distinction: the number that’s the lowest isn’t necessarily going to be the number with the most negative momentum behind it. For example, let’s say you’re a martial arts school. You have 3 basic key metrics: students who sign up, students who come in for a trial class, and people in your audience. If you look at your data and see low signups with a flatlined trend and trial class appointments with a negative trend, which is more important?
The average sales and marketing person would say, hey, sales are low! We need to focus on closing more sales! But this would be incorrect – the most negative trend is the fact that you’re losing trial appointments. That negative momentum is the battleship you need to steer first, and once you fix that, all of the boats tied to it will start to change as well.
Start with the KPI that has the worst trend and fix from there.
Tomorrow the blog will be off for the 4th of July, and Thursday we’ll look at something else from the finance world as applied to digital marketing.
You might also enjoy:
- Is Social Listening Useful?
- The Biggest Mistake in Marketing Data
- Retiring Old Email Marketing Strategies
- Marketing Data Science: Introduction to Data Blending
- How To Start Your Public Speaking Career
Want to read more like this from Christopher Penn? Get updates here:
 Get your copy of AI For Marketers |
Great post, I could not agree more that trends are important. But another good way to add context is to look at year-over-year trends. While a month-to-month trend may show downward movement, the year-over-year trend may actually be increasing. This is often due to seasonality.
Best,
Justin
In a crisis- it’s interesting that people focus so much on symptoms (sales dropping) instead of trying to identify the real cause(s) of the issue. Have you ever applied the “Satisfaction Triangle” (Process, People, Product) to your studies of marketing?
Chris, thanks for this post! Recently, we have added new functionality to the Shopping Cart and since we didn’t do A/B test, it wasn’t very effective to compare cart conversion before/after. With moving averages, we can spot a trend much easier (I found a post about adding 10-period MA and 30-period exponential MA — wherever they cross, it is a trend change). Good stuff!