The Criticality of Great Analytics

“Good data in, good data out.”

Data and reports should be actionable.

Clients don’t want to pay for reporting.

I’ve heard these sayings thousands of times in agencies and brands around the world over the last 20 years. So why is it that so few agencies are able (or willing) to invest in true data, true science and true insights in digital and social media? Sure, part of this is because clients don’t feel they want to pay for, well, anything, but if we’re honest, clients will pay for things they personally find valuable.

So, maybe the question is: how can we make our data/analytics/insights so valuable that clients can’t live without them?

There are four key things we must fix in order to get our clients this type of epic win value:

  1. Our actionable data isn’t actionable
  2. Our data points must go deeper
  3. We must create a feedback loop
  4. Balancing the reporting value equation (more on this later, I promise)

A Quick Pause

I will note that this post isn’t a slight in any way, as there are agencies that take a value driven approach to analytics and reporting. In addition, this post will highlight my beliefs on this, so if you disagree or have a different approach, I can’t wait to hear from you!

Making Things Actually Actionable

Often when we say we want our reporting to be actionable, we’re lying to ourselves. When was the last time you did an action based on a 3% increase in pageviews? How about based on if you got 130 or 100 comments? Have you ever made a decision based on your Facebook engagement rate? When was the last time you increased or decreased a media spend based on actual performance?

So, how do we fix this? We need to start with a blank slate or white board and write this at the very top:

What are the questions the business need to answer in order to grow over this report’s time period?

After all, how can we action something that doesn’t impact the business? Clients simply won’t care. By focusing on the specific questions we need to answer in order to grow the business, we can begin to construct appropriate data points. Some examples of questions are below:

  1. Which narratives are driving value?
  2. Where are we getting the most bang for our buck?
  3. What do our users care about?

These are more esoteric questions than most reports are built around, and how you answer them will be different for each client. So let’s look at an example: “where are we getting the most bang for our buck”. In order to answer this, we need to measure a lot of things, like:

  • Baseline performance
  • Impact of media (earned and paid) on performance
  • Competitor performance
  • Platform changes
  • Etc

Ideally we’d measure other things like Topical Engagement Ratio (General engagement vs category engagement), etc, in order to get to the answer to this question, but it’s hugely actionable: it allows us to invest more in content or channels or platforms or narratives or tactics that work, while deprioritizing those that don’t. If we can collect smart data, we can increase value for our clients without having to increase budgets. In addition, being smart for our clients can create a significant value moat vs other agencies by making our clients smarter.

Deeper Data Means Deeper Answers

Alright, if we are using data to get to answers, we need to start to ask deeper questions. As an example, if we collect data on impressions… is impressions for a post really that valuable? No. You know what might be valuable? Velocity of traffic: when and where awareness increases and decreases and why.

How about:

  • Not how many replies or RTs or Pins we get, but how much downstream content we create, and which nodes in the graph are creating the most awareness (aka: we can then engage them directly)?
  • Not how many comments you get, but who are the most engaged users in your content and what is their individual sentiment / affinity to the brand?
  • Not top posts or categories, but how a single narrative is impacting the business across all touchpoints?

By looking deeper at our data, we get deeper answers and we can often turn “dumb data” (aka: pageviews, Likes, etc) into Smart Science.

Building a Feedback Loop

So, if you are:

  1. Asking smart questions; and
  2. Using deeper metrics to generate smart answers

Then, the next step is to build out your content creation, strategy, creative and account processes to include the specific learnings from these reports. I mean, I know we tell our clients we optimize based on these data but, if we aren’t actually doing this, we’re producing reports just to make ourselves feel good. It’s the Digital Strategy equivalent to expensive toilets that the Pentagon keeps buying (or would if it was running right now…): a whole lotta effort for a whole lotta crap.

On that whiteboard where you were writing your questions? Write how you’re going to actually action and build those actions into your feedback loop. Any actionable insight that you don’t push into a feedback loop is one you never action on.

Smart Reporting Is Based on Timeliness

Reporting Timeliness

Reporting Timeliness

I use this ugly graphic a lot, because it’s how I visualize this critical issue.

To recap, if we are making our questions/answers actionable, based on real data and integrating it back into our process there’s one likely issue: our reports will be fucking massive. And not in a good way. After all, very few clients will let us spend 100 hours on a weekly report (nor would most sane people want to!).

Our reporting infrastructure and process needs to trade the value of a data point with the timeliness of it. There are certain data points that are inherently valuable in real time, like that a competitor is kicking off a media campaign. Then there are certain data points that are inherently more valuable by taking a wider view, like platform performance or competitor whitespace… which are likely better for quarterly or annual reporting.

While the right reporting process is different for every client, many of my clients have tended to fall into the following buckets in terms of reports:

  1. Weekly Campaign: Snapshot Dashboard
  2. Monthly Report: Digital Dashboard, Campaign Performance, Topical Engagement Ratio, Key Competitive Analytics, Narrative Performance
  3. Quarterly Report: Channel Review, Competitive Insights, Content Planning Analysis, Funnel Analysis

Of course, in an ideal world, we’d all have digital dashboards which could allow the customization of these data points, and generation of many of these reports automagically (with Insights to be added), but that’s a future post. These are magic, if your agency can find a way to justify them, as it means you’re spending your time on value and insights instead of gathering dumb info.


I will be working on a deck for this over next couple of weeks, to show examples of these types of reports and how to gather them, but if you only take a few things away from this post, I hope the following ring true:

  1. Good data means deeper data.
  2. We need to make actionable data actually actionable.
  3. Smart reporting balances timeliness and value, to allow us to actually optimize.

Oh, and, yes, clients will pay for where they see value, so you get to stop doing the “dumb work”, and start becoming more of a “smart partner”. When is that not a win?

Argue and ask questions in the comment section belowww.


8 thoughts on “The Criticality of Great Analytics

  1. Hmmm nice post. I like your point on timeliness… often times there is a focus on ‘real time’ insights. Information should be as timely as how quick a decision can/needs to be made. Example: If you can’t change the course of action in less than a 1 week span, you don’t need a daily report.

    You touch on it here also but the one thing I would add is the value of reporting vs. analysis. Reporting is the plain statement of numbers which often times can be overwhelming especially when it’s unfocused and spanning over several different metrics. Where the actual value in reporting is when everything is synthesized and and communicated in a intelligible way (ie: analyzed). Purely looking at numbers can be dangerous as we often get obsessed when things go up or down, and not how they trend and create impact over time.

    PS: That picture looks familiar…..

    • Jeremy Wright says:

      Hey Kev, thanks for being the first commenter ever on this project!

      I totally agree that consolidation, analysis, insights and visualizations are absolutely key to ensuring that your Science and intelligence is actually digestible by the client. Obviously step one is getting the science right, which was where this post is meant to start a chain of posts 🙂

      The pic looks familiar? Unless you’ve seen one of my decks in the last 18 months, I can’t imagine why…?


  2. Great new blog. I will be reading with interest. Our radio clients in particular have a real issue with #1 Our Actionable Data isn’t Actionable. Their problem isn’t so much doing poor digital strategy but even understanding digital strategy. If reports aren’t completely actionable in red light – green light formation, then reports might as well not exist. Congrats, look forward to discussions.

  3. Jeremy Wright says:

    Elliott, always great to see your name. The integrated reporting challenge is a distinct issue I hope to cover at some point. Not an easy solve but one that clients (rightly) demand.

    Plus. Always happy to grab a beer mate hah. Looking forward to your thoughts on upcoming rsnts!

  4. Pingback: Getting to Magic: Facebook’s Negative Feedback Data | Strategerize (BETA) - One Man's Quest for Better Strategy

  5. Pingback: Front-End vs Back-End Reporting | Strategerize (BETA) - One Man's Quest for Better Strategy

Educate Us - Leave a Comment!

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s