5 ways to use context to find hidden insights from your data.

Updated: Sep 15, 2020

Data is Everywhere. We marketers, analysts, and businesses love to collect data as much as we could. As someone visits our website or our Competitor's website (no we don't want that!) to buy something we collect data about what products have been viewed? how many times customers come back before they make the purchase? where they came from, what products have been bought? Their color, size etc. And now with the rise of smartphones, we practically become the data Gods. We now live and sleep with Data :)

Which leads to a big problem for us also. Finding insights from this vast amount of Data. Now we do not only have to be more strategic but also clever when dealing with data. And one of the greatest friends in our data journey is 'Context'. Without it, we all will be lonely and miserable.

Context helps us to understand what is good, bad or ugly. When we apply context to our data, everything starts to make sense which wasn't obvious before. It helps us to focus on things that important to us rather than chasing the rabbits all over the places.

In this post, I am not only going to show you 1 or 2 but 5 ways you can use context to find hidden insights from your data and take your analytics skills to the next level and most importantly growth and success for your business.

1. Comparing trends over different time periods -

Let's say that you just joined a company in January and don't have any ideas about your companies marketing strategies or anything else. And You opened the Google Analytics reports and you see this.

 In this case, I am looking from Jan - Dec 2017 data. You can see that after July there is an amazing upward trend to the right, which is a good sign as you are getting lots of more traffic.

And in the second pic, you can see data like New users, sessions, pages/sessions, bounce rate etc.

Everything is looking nice and Good. But the problem is it doesn't tell me anything interesting that will make your boss pat on your back.

Now, Pic any comparable time period, this month to previous month or previous month last year or maybe this quarter to the previous quarter. Pick any.

In this case, I am comparing 2017 to 2016.

Now Look at the below pics, did you find something interesting?

If you look at the above pic, you can see that the number of users has decreased, the number of new users, sessions, pages/sessions, pageviews all has decreased during 2017 compared to 2016.

Holy Cow! Somebody's is going to get fired :)

Wait, don't worry.

Look at those two green signals, Average session duration and the number of sessions per users both have actually increased, which is a good sign that this website actually gets more qualified engaged traffic than before.

Let's dig a little bit deeper to understand what outcomes were delivered during this time period. First, go to the Conversions > Ecommerce> overview.

Here is the complete picture. Though traffic has gone down but Revenue has actually got doubled from $2 million to $5 million. wow, That's freaking amazing.

Ecommerce conversion rate, transactions, average order value all has gone up.

And if you scroll down a little bit in the same report, you will find these numbers. Though the number of transactions has decreased both for campaigns and order coupon code but both the revenue and average order value has increased, means more high-value users and fewer opportunity seekers. Look at the amazing 300% increase in revenue and 850% increase in avg. order value for coupon codes.


Everything has increased significantly, so the marketer doesn't need to get fired. Actually, he/she deserved to be promoted and should be given 2 weeks of long holiday in the Maldives or Santorini Greece :)

See how much clearer it becomes when you apply a simple context to your data. Everything started to speak for itself, loudly and clearly.

So now, Go and find out, what the marketers, other team members did to get these amazing results. What kind of promotions, campaigns they ran or stopped to achieve these results. These will give more context to your data and you can do more of it what is already working for you, and stop all those activities that aren't delivering any high-value.

want more awesomeness?

why stop here?

Now the Analyst also want a vacation in Hawaii :)

How will he do that?

He will leverage the power of segmentation.

2. Using Advance Segments -

If you look at the Revenue trend for 2017, you can see the same thing as we seen before, sometimes it's up and down, an only interesting thing in this data is revenue has increased to the highest level during the Nov and Dec. Nothing much to get from here.

Now let's apply segmentation to this report. Pick any of your most important traffic, campaigns, medium or anything else that important for you and create a segment of it.

In this case, I am just picking few acquisition segments where this website is getting most of the traffic - Referral, organic, direct and social.

Now, look at this. Do I need to say anymore?

The Revenue from the referral traffic was below the direct traffic during Jan to Mar but it starts to keep getting better and better after April till Dec.

And at the same time, the revenue from the direct traffic plummeted to a very low level.

What happened there?

The customers who came from direct traffic usually are the people who know your brand well or find it valuable and took the pain of remembering it or bookmarked it to later come back again.

Then why this traffic is performing so worse?

Which product pages they are looking at? What they are doing once they visit your website? Aren't they able to find what they are searching for? And why isn't anyone inside the company is trying to find out what is actually going on? Lots of questions need to be answered.

Second, who are these lovely people, who are coming from the referral? Which websites are sending these high-value customers to your site? Don't wait, just go and find out. Give them a big hug or maybe a kiss :) for sending these amazing people. Make a better partnership with them to increase more sales and revenue.

Simple fundamentals but deeper insights for you.

3. Industry Benchmark -

Another great way to give context to your data to understand your performance is Industry benchmark data. It will help you to understand how is your business doing compared to other business in your industry.

And you don't need to go or beg anyone for this. Just look at your Benchmark report in Google Analytics. Benchmarking Report in Google Analytics -

First, go to Audience > Benchmarking > channel

As soon as you launch the Benchmarking report in Google Analytics, It will automatically choose the industry verticle and size of your business and compare it to other websites with the same size as yours in your industry. It will also show you how much websites are contributing to this benchmarking report ( Look at the top right in the first pic.)

For some reasons, if Google analytics doesn't do it automatically for you, then just set it by yourself.

To choose the size by daily sessions - First, go to Audience > overview

Then pick a shorter time range, say 1 month and change the metric at the top left from users to sessions( Shown by arrow). Then look at it by day( also shown). You can see from the above pic that this website is getting less than 5000 sessions on an average per day.  That's the range you will set in the Size by daily sessions drop down menu.

But before we discuss benchmarking reports settings and how to use it. First, let's understand why this is a good source of data to get some context. Why Is it a Good source of data?

Because the data in this benchmarking report comes directly from all the websites that are using Google analytics, like yours and mine and choose to share it anonymously with Google. So this is a very pure source of Benchmarking data that you can get to measure your performance with other businesses in your industry. Another great thing is that even if you have a very small number of sessions on your website, you can still benchmark your performance which is very awesome.

Benchmarking report settings -

1. Industry vertical ( Required) - There are more than 1600 industry verticals and sub-verticals so you can be very specific to define the industry your business is in.

2. Geographic location (optional) - If you wish to limit the benchmarking data to a specific country, you can do it from here or maybe you want to measure your performance against any businesses in other countries to get a sense of competitiveness you can also do that.

3. Size by daily visits ( required) - Select from seven traffic size classifications. This setting allows you to compare your website against websites with similar traffic level in your industry.

you can also choose any other metrics that are relevant to you and compare it with other metrics, which will be shown on the trend graph ( Below pic).

Without any further delay, let's dive into the Benchmarking report.

It's obvious from the above pic is that this website is totally crushing the benchmark. Also, there is a huge spike on 12 &13 Dec, finding more about it can help you to repeat this kind of similar results. Whenever you find something like this, go down the rabbit hole and find out the reasons behind it.

In the Above report, there are total 6 metrics available and divided into Acquisition and behavior category, 3 in each category. Another thing is that this data is already segmented by traffic channels so you don't have to do anything to understand your benchmarking performance for these particular channels.

At the top, you can see the overall performance of this website. You can see that in the Acquisition section, sessions, % of new sessions, and New users all are performing better than the benchmark data.

But if you look at the Behavior section, you can see that all the 3 metrics Pages/sessions, Avg. sessions duration and bounce rate are all performing worse than the benchmark. Which is a direct indication is that this website is very good at attracting customers but very bad at engaging with them. Which are a very bad sign and they need to figure out some ways to engage with the customers more.

Looking at the below channels can help you to get an idea, which channels are bringing low engaging traffic and then you can make a plan to improve its performance.

Let's clean the report a little bit to analyze the data more effectively. There is a way to much stuff going on here.

Much better than before.

Look at the two little buttons at the top left. The first button will remove the raw numbers, which I just did. And the second will remove the heatmap but I like it in my report. Do according to your preferences.

If you look closely at the above pic, you can see that Referral traffic is doing so much better. Not only it is good at attracting more customers but these customers are also engaged with the website. Learning more from these customers can also help you to improve other channels.

Apart from referral if you glance at the report, you can see that most of the high traffic channels are good at acquiring customers but very poor in engaging with them. The report is totally filled with red color in the behavior section. And the last 4 channels isn't good at anything.

Looking at this report can help you to get a lot of ideas about which channels you need to improve and on which channels you need to double down your money and other resources.

Isn't that Amazing?

Wait there is more.

Go to the Benchmarking > Location,   just below the channels report in Benchmarking section.

Nice amazing benchmarking report segmented by country.

Here you can see that the website is outperforming the benchmark in the United States. It is doing good overall in both in acquiring and engaging the customers in united states. It might be that most of the referral traffic is coming from the United States as we saw above that referral traffic is also exhibiting this kind of same behavior and also the revenue and transaction are higher for referral traffic. Segmenting your All Traffic > channel report in Acquisition section in the left navigation in GA by Country and help you validate this hypothesis and also help you to get additional insights from it. Always segment your reports, never use a report without segmentation.

Apart from the United States, this website is performing below the benchmark in other countries. Same here, good at acquiring traffic but very bad at engaging with them.

Let's move on to the Device Benchmarking report.

The % of new sessions and New users are all outperforming the benchmark in all the 3 device category but the number of sessions is underperforming on tablet and mobile device. This means that there might be some issues with user experience on mobile and tablet devices. Customers aren't using mobile and tablet devices much as their preferred channel for interacting with the website.

And if you look at the behavior section, you can see the same thing as we seen before, the behavior metrics are underperforming below the benchmark. As you already saw that this pattern keeps repeating again and again. There is some serious need to solve this issue. One of the best way to find out why and how to solve it is using qualitative data, which we are going to talk about later in this post.

4. Competitive Intelligence data -

The reasons you should use CI data to get some context is that It clears the fog ahead of yours and your management's eyes. It shows you the reality of the world. Let say you are getting 100K monthly visits to your website or your website is constantly showing up on google search result for your most important category terms in your industry but the question is how do you know what is good and bad? How do you know that a 100k monthly visit is a good number? What if your competitors are getting more than 200k or 500k monthly visits? what if the share of search for your category terms are only 10% and your competitors are getting more than this or worse somebody else is stealing the traffic from these keywords whom you never heard off. A total nightmare!

Right now in India, there is going a battle of giants between Amazon and Flipkart. Two biggest E-commerce company in India. Let's see how getting some context from CI data can help you to understand your performance and also find new growth opportunities.

If you look at the above pic, you can see that during Dec 2017, Flipkart gained the highest amount of traffic to their website during this 3-month time frame. You are happy that you break the barrier of  250M visits per month and planning for a big party after 6 pm. Then someone came and show you this below pic!

Now, did you see the problem if you don't apply context to your data? Amazon is totally crushing Flipkart. Not 1M or 2 or 3M, they are getting more than double the amount of traffic that Flipkart is getting. So how does it a victory for Flipkart? Now Flipkart need to bust their A%% and have to figure out, why is this happening? How did we end up here? How we can beat Amazon? We bring the E-commerce revolution in India then how did Amazon get ahead of us? What are they doing better than us? One simple context raised the whole new set of questions and priorities inside the company.

Now you know, you need to fix this. First, go to the Channel overview report in Similarweb.

From the above report, you can see that except for Display ads, Amazon is beating Flipkart in each and every major marketing channels.

Now, you know that you need to work on your Organic and paid search strategies. you need to improve your referral and direct traffic etc. So now, you can go to the organic and paid search report in Similarweb and find out which keywords are sending the most traffic to Amazon and create a better paid and organic search strategies to get ahead of them. Look at the referral traffic report to find out which websites are sending more traffic to Amazon and create some kind of partnership with them.

And if you are a marketer in Amazon then you should dig into the Display traffic report.

Look at all these amazing data above. From the Ad creative report, you can learn what they are so good that they are getting so much more traffic than us. you can look at the websites where their ads are being shown. You can look at Ad network and publishers report to find new opportunities for your own growth. Learn what your competitors are doing good then beat them at their own game.

There are lots of amazing data in Similarweb. Totally leverage it to understand your performance compared to your competitors and then plan better strategies to steal traffic from your competitors. Totally crush them as much as you can.

To learn how to do competitor analysis and find insights for your own growth, please read this post - Competitor analysis of Zara, H&M, and Topshop with Similarweb and Google trends.

5. Qualitative data -

One of the biggest problems with any web analytics tools is that they are very good at telling you things like How many customers visited your website? What products they look at or purchased? How many customers added products to their cart but didn't make the purchase? blah blah as we discussed earlier.  But they never going to tell you why in the first place a customer chooses to buy a product from you? There are many other options but why you? If they made up their mind to make a purchase then why did they decided to abandon it? The big "Why" part. Pick any web analytics tool, anyone you want, they are very bad at telling us this "Why". Why someone did what they did?

You can run as much A/B test or Multivariate test as much as you want. But you only get the correlation between customers intent and behavior but not the causality. And we all know that correlation isn't equal to causality.

To get that "Why", you need to add some qualitative data with your quantitative data. Then you will say something like " Oh I see, this is actually the case. Now I get that."

Let me give you not one but two examples of getting context from qualitative data.

if you look at the above pic, you can see for queries related to Competitor analysis, my article is appearing on the number one page of Google organic search result. From this data, I know that how many impressions I got, how many people clicked on the article to read it on my blog and in which position my article is appearing on the search result page?

But none of these data is telling me Who are these readers? Am I able to help them in getting the job done for which they came to my website? If not then what was the reason for this? Am I didn't add the relevant things in the article? or is there something else?

Now, let's add a little bit of qualitative data to it.

Far much better than before. Right?

Now, I know who they are? In this case, some of them are University students and they are reading the competitor analysis post for some project in their university. Some people came to learn and download the remarketing list. From the above data, Now, I know what was the purpose of their visit to my website. And With the second question- " were you able to complete your task today? I am adding another level of context to it. Some of them say yes and No. If No, then why not? With the 3rd one, I get that. It's not Inception, it's Contextion :)

One more, why to settle for less :)

As I wrote mostly about web analytics and digital marketing ( Soon thinking about expanding to more things.). Anybody can understand that my readers come to learn about these topics.

But, I wanted to go one step further, try to understand what was the end goal? What progress are they trying to make in their life by reading and learning more about analytics and digital marketing?

Now, I have a better understanding of it. Someone of them trying to grow their online business or get more traffic to their website or becoming more confident in the analyzing data. I am sure some of you might have some other reasons. But Now, I have a better understanding of some of my readers, what actually they are trying to do, better than the number of pageviews or sessions etc. I also know, some of the things they are struggling most to get that job done.

Not these insights helping me to know my readers much better but if you look closely you can see that these insights are also guiding me what content I should write more about. Isn't that just amazing?

You will never able to get these kinds of deeper insights from your web analytics data. And all the strategies outlined in this article is about that, finding better insights and opportunities to grow your business.

  • LinkedIn Social Icon
  • RSS Social Icon


Life With Data