How to make the most of your data analysis
This is a common question when you have a lot of data to look through. A common answer to this is to start with a list of questions you would like to answer when you’re working on your data analysis. Simple questions like: “how do my customers interact with my site?”, “what is my best selling product?”, and “how long is the customer journey from first click to sale?”
This approach is fine, and you’ll no-doubt find the answers you’re looking for, assuming you have the raw data available and you’re able to segment it effectively. However, I believe there is a better way.
Leave your assumptions at the door
There is something called a “Scotoma” which is (in short), a mental blind spot.
Have you ever had that conversation with your partner? It goes a little something like this:
Partner: “Have you seen the rice?”
You: “Yes it’s in the cupboard”
Partner: “I can’t believe we don’t have any rice!”
You: “We DO have rice, it’s in the cupboard”
Partner: “We must have run out, it’s not here – I can’t believe we don’t have any rice”
This conversation could easily go on for a decade if you’re not careful, so you promptly go over to the cupboard, pull out the rice, slam it on the worktop, and give a meaningful look to your partner which says “open your eyes!”
This is a common occurrence in all walks of life. A scotoma is a mental blind spot built up by belief. In the example above, your partner’s belief that there was no rice in the house literally stopped them from seeing what was clearly right in front of them. It just didn’t register.
The same is true with data.
Imagine you’re a small business owner, you spend all day talking to your customers on the phone when they get in touch and feel you have your finger on the pulse. You go to a marketing agency and the conversation goes a little something like this:
Business owner: <says with confidence> “Our customer base is 70% male so we should be focusing our marketing on that”
Data Analyst: <looks at the data based on website visits, time on site, engagement with content> “No, your customer base is actually an 85% Female audience in terms of your website”
Business owner: “No, no, no. I speak to customers every day – I know my customers. 70% men. As I said”
Data Analyst: “No. They’re not – look at the data”
Business owner: “I know my audience”
Data Analyst: <spends hours constructing a presentation to demonstrate how a customer journey works>
Business owner: “Hmm… but still…”
Imagine if the business owner insisted on running the marketing based on their assumptions? Ads would be tailored accordingly, audiences chosen, money spent – all in the wrong direction. The result is either a failed marketing plan or a mediocre one at best – and this is the simplest of examples.
We are all guilty of making assumptions. We all approach our businesses with certain bias. Its natural. The problem with this is that your assumptions will blind you and could in the long term, damage your business. You need to let go of the assumptions. Use data analysis to get past your scotoma and open your eyes.
Let data analysis guide your questions
Some people misunderstand marketing data analysis. There is a belief that you can glean all the answers from “the data”, which, to an extent, you can if you have the right tools and the time available. But to get beyond the simple stuff, you need an enquiring mind. You need to be looking at the data and constantly asking “why?”
You need to find the basics within the data first, answer your initial questions if you can, but then allow the answers to your most simple questions guide your thinking and lead you to further, more in-depth questions. Ask yourself frequently “what if I’m wrong?”, check, and re-check the data.
Re-cut the data with a specific question in mind now that you have the information to hand. Then look at the problem from different angles with different tools.
For example, you’re surprised to find your average order value one month is abnormal, its higher than usual but you’re not sure why.
You look into your raw sales data with this question in mind and find that there was a spike in sales on a single day during the month, for a single product with a higher than normal average order value. It spiked your average order value overall for the month but that product’s sales were lost in “bigger picture” when looking at the month without any analysis.
But why did that happen?
You look into your traffic and find a similar spike in organic traffic on that day.
So you check your search console and keyword tracking data and find that a keyword you hadn’t considered optimising for had jumped up on to page 1.
So why did the sales stop?
You note that the keyword isn’t ranking on page 1 any more. It must have been a blip with the Google algorithm, but now that you’re paying attention you realise that you have been ranking for that keyword on page 2 for months.
So what now?
That keyword has already proved your case. It generated sales, converted well and increased your average order value – all in one go. Now you need to dedicate the time to optimising for that keyword to secure a more permanent page one placement.
Why we love data analysis for marketing
Making assumptions is normal (and hard to avoid), even when we know we shouldn’t be making assumptions at all.
What we love most about marketing data analysis is that if you allow the data to do its job, use the right tools, assess objectively and ask the right questions, the answers can completely break your assumptions about the business, their customers and the objectives you should be prioritising. It’s an education process and a constant evolution which benefits everyone involved.