Context is everything.
Sure, it’s an old cliche. But it’s more relevant than ever when it comes to data. In fact, we’d go as far as to say that data’s pretty useless without it.
OK, hear us out.
Data is super valuable; it drives decision-making. But it can’t do all the heavy lifting alone.
Think about a mechanic looking at a car with an electrical fault. Today, these are loaded with more tech than ever so knowing there’s a problem isn’t enough.
When was the car last serviced? Has the software been updated and if so, what version is it on? To get anywhere fast, this mechanic needs to know the context.
How about this: you’re heading to the shops and it’s a 30-minute walk. You could either take a leisurely stroll or jump on the bus. Well without even considering things for long, several circumstances could affect your decision:
- What’s the weather like? Is it worth taking a scenic walk to get there?
- Do you have the money for public transport?
- Do you need to get some more exercise that day?
- What’s your timeframe? Are you in a hurry?
And this is just one small decision you make; understanding how your customers think and feel is even more complex. Numbers just can’t tell you the full story.
Using context to get a clearer picture from web data
Without the proper context, web data can be easily misinterpreted. Worse still, it can also be skewed in order to tell a story that’s somewhat misleading. For example, just take these two tried-and-tested narratives we've grown to accept over time:
Dwelling is good - it’s been drilled into us that a good average session lasts between 2 - 3 minutes. But doesn't it depend on the website? Sure, if it’s content-rich with lots to see and do - you want your users to stick around. But if you’re quickly trying to direct a visitor to sign up or purchase on a much snappier landing page, maybe not. Who’s visiting the site? What are they here for? Is the navigation overly complex or confusing?
Bouncing is bad - again, this one comes down to what the visitor’s trying to do and where they are on the site. Alarm bells usually ring when you see a high bounce rate - but sometimes they’re perfectly normal. For example, if someone is skim-reading a blog, they don't usually dwell for long, On the flip side, high bounce rates and low conversion rates are always bad - so focus on those instead.
Finding the throughline between different data sets
OK, it’s true that the better quality data you put in, the greater the insights you’ll get out.
Just don't forget about context, though. That’s how you read between the lines of your datasets. Here’s a few ways to ensure you tell the right take (and keep that story straight):
Don’t overly on a single data source - instead, connect multiple data sets (both qualitative and quantitive) with contextual circumstances to get the best view.
Tell it like it is (even if it hurts) - when it comes to graphs, don’t play around with the axis scales or reporting durations. We know it's tempting to get the best-looking picture - but it can trip you up in the long run when you need accurate insights. Instead, tell your tale with concrete facts - not convenient fabrications.
Let it do its thing - finally, try and be patient and don’t overreport on data. A good tip to remember is that if you can’t affect change within a day, don’t check it daily. There’s just no point in wasting your time.