Naming the Data We Use

Quote from Wizard of Earthsea

With starting over again in the freelance world and launching the new business/website, I have become conscious of language. I need to use the language of my readers, prospects and (hopefully) customers rather than my own language, as that may be jargon.

This is important to me as I expand the scope of the data I use from predominately Google Analytics (GA) through to all data available to online organisations. I want to ensure I use a term potential clients recognise when I refer to GA type data (in case they use an alternative like Adobe Analytics (AA)) and the same for when I describe the data I now work with.

My desire for these definitions is influenced by reading about product positioning and thinking through how that relates to me. In part, I think if I can get these terms right, it does position the work I did in the past and will do in the future. But it is more the desire to use a term so that people instantly get what I am referring to (even if it is not technically correct).

To clarify the language and correct terms, I put a couple of simple polls out on Twitter and LinkedIn. The first asked for the term used for data collected from tools such as Google Analytics and Adobe Analytics. The second poll asked for the term used for all data collected from multiple sources for online organisations.

Tweet asking poll questions

For reference, I currently refer to the data from GA/AA as Digital Analytics data. I have been wanting to use the term Digital data to include data from all data sources for online organisations, with a second preference of Business data. But I have been confusing myself on how to define each term which suggests issues for anyone reading.

LinkedIn vs Twitter

Quick side point on the two platforms. The polls were part of the same thread on Twitter and were open for 10 hours. They were separate posts on LinkedIn and remained open for 3 days, as I thought LinkedIn might keep them active in notifications for longer. In hindsight, wish I had kept the polls open for the same time on both platforms as an experiment.

Poll results from Twitter
The first poll had 600 impressions on twitter with 33 votes while the second had 559 impressions and 26 votes. As they were part of the same thread, I was expecting similar numbers for each. LinkedIn clearly outperformed Twitter with 7,490 impressions for the first poll with 184 votes while the second poll had a lesser 3,277 impressions and 60 votes. You can’t include two polls in a single post on LinkedIn which could explain that different behaviour across the two platforms. There were only a couple of replies on twitter but many comments on LinkedIn, particularly for the first poll.

If I had predicted in advance, I would have guessed at more votes on Twitter.  Having more comments on LinkedIn was less surprising. It would be interesting to learn more on tactics for different platforms for the future.

Results of the Polls

Thank you to everyone who responded and voted in these polls. Slightly to my surprise, there was no overwhelming consensus.

Poll results from LinkedIn

According to LinkedIn, the correct term for data from tools such as GA or AA is Digital Analytics. Frustratingly, the correct term for all data for online organisations is apparently also Digital Analytics. With a smaller sample, Twitter agreed on the first poll verdict but with Business Data as the top results for all data sources.

As was pointed out, I missed a common name for Google/Adobe Analytics type tools – that of Web Analytics (I had meant to include this but forgot it when typing up the polls). Our community, led by the Digital Analytics Association, rebranded from Web Analytics to Digital Analytics nearly 10 years ago. This was to reflect that data was coming from more than just website analytics tools (which is similar to my question).

With thanks to Kyle Akerman for the suggestion, we can see that while Digital Analytics is trending up over time, it still lags behind the term Web Analytics which has been trending down. While the rebrand may reflect the data sources, it doesn’t necessarily reflect standard usages of the terms.

I now wonder if this rebrand confused me and other analysts. It led to tools such as GA/AA being referred to as Digital Analytics with the output being Digital Analytics data. Which doesn’t leave any space to move when referring to this data plus data from other online sources. Hence how the LinkedIn polls produced the same result for different questions.

An interesting alternative term used for the output from website tracking tools is Behavioural data. If you can read my comment responses chronologically, you will see me say no, no, hmm maybe when this is offered as the alternative. One of the people who proposed this name was Alex Dean from Snowplow so it was interesting to see how the tools describe what they do.

Terms used by Vendors

Snowplow reflects Alex’s comments, referring to Behavioural data as per the screenshot below. Google Analytics doesn’t have a marketing page directly (you redirect to the platform) but interesting to note their snippet on Google appears to be slightly out of date by referring to tracking Flash. Via Google Marketing Platform, there is more copy referring to customers and understanding how users engage with your website and app – although without the word “behaviour”.

Adobe Analytics appears to use the term web analytics and web data, allowing you to bring in marketing data. Webtrends goes with Digital Analytics while AT Internet is more Web Analytics. Matomo is just “not Google Analytics”.

What I found most interesting there is the lack of a common name or description across these tools. They have thousands of clients each and I assume a fairly large history of research into using the right language. Which means there is no clear consensus from the poll responses or across the vendors.

What is Analytics Data?

On a side point, I find myself going around in circles with the phase “analytics data”. What does that even mean? Is there “non-analytics data” or “analytics non-data”? Should the word “analytics” be dropped leaving us with just “data” (of different types).

The work on the data though is performed by Analysts. The data is provided by Analytics tools. It feels natural then to say it is data from Analyics tools and therefore Analytics data, meaning I am suddenly back where I started…

My New Definitions

I think I have a simple answer for my definitions. GA/AA type tools can revert back to being referred to as Website (or App as applicable) Analytics tools. It is when this data is reviewed in combination with other data sources that it becomes Digital Analytics.

This results in a simple approach:

  • Website/App data – is data on websites/apps from tools like GA/AA
  • Customer data – is data about customers, available in CRM tools (or anywhere else)
  • Marketing data – is data from marketing tools (or related to pre website spend/behaviour)
  • Digital data – business data related to online performance
  • Business data (or data) – covers all data available in a business (including salary, rent, etc)

For all of these, the word “analytics” can be inserted prior to “data”, if you so choose (so marketing data becomes marketing analytics data, etc). But I think “analytics” should refer to the work we do and “data” to the data we work with.

The integration of other data sources into Website/App Analytics tools does convert them into Digital Analytics tools, equivalent to a Data Warehouse that has that same integration of multiple data sources. But I question how many companies are truly operating with that integration within their Web Analytics tool. If they are at the level of data maturity where these data source integrations are managed, then it is most likely to be combined outside of the tool e.g. using Google BigQuery, Adobe Data Warehouse, Amazon Redshift, etc.

Reviewing that search terms data, we can see that both Digital Data and Website Data are more popular terms than either of Web Analytics/Digital Analytics, at least in terms of search volume and recent growth. This suggests I am on the right path with my approach.

A Future Survey

When I created the polls, I wanted to know what businesses thought, not analysts. I have a strong suspicion my data was skewed by analysts responding with the language they use (and believe businesses use). If I repeat the exercise, it will be a survey including the question on job role, so I can segment the results.

Behavioural Data

I am still adjusting to, and deciding how much I like, the term behavioural data for data from GA/AA/etc. While it feels accurate to describe data collected on visitor behaviour on a website as behavioural data, I think the Website Analytics tool provides more data than this and therefore the term is not comprehensive enough. It could also lead to confusion as behavioural data is used elsewhere e.g. to describe the output from focus groups.

I can see behaviour data being the right term when used to discuss the data collection that Snowplow does, so it is great for positioning them. But it is not a term I am going to use myself as it may be technically correct but I am not convinced that it is a term my potential clients will instantly understand.

Feel free to disagree with any or all of the above in the comments…