How to implement a data-driven media budget


Check out all the on-demand sessions from the Intelligent Security Summit here.


Data is king. It was a fact several years ago and it stands as a fact nowadays. For marketers, the problem is not to dive into all data; the key is trustworthy data. The dilemma is that, while having data that is measurable and useable, marketers must also accept that true measurement won’t be coming from single source data points.

We have an advertising landscape made out of several walled gardens. These are inventory suppliers, such as Meta, Google and TikTok, that do not provide open access to 3rd parties. So what can marketers do when the wrong decisions can be costly? 

Traditional media such as linear TV, out-of-home (OOH), print and radio are not yet players of the past — they are still some of the biggest suppliers. And with good reason — return on investment (ROI) estimates for linear TV are still better — 18%, 22% or 25% lift on digital ROI, depending on the source.

The problem always arises when we look too closely. For instance, Kellogg Insight found contradictory insights to the above, claiming that TV Advertising is not usually worth it.

Event

Intelligent Security Summit On-Demand

Learn the critical role of AI & ML in cybersecurity and industry specific case studies. Watch on-demand sessions today.


Watch Here

So, who do we trust? And how much can we trust their data? This article is not about TV, but we’re using TV to make a point — even the most widely used and researched inventory in the world will produce conflicting results.

So, what are the key steps to approach data measurement? And, how can marketers correctly collect and leverage data to optimize media budgets? 

Firstly, recognize the problem

Untrustworthy data is usually due to a couple of factors:

  • Attribution is an unsolved problem
    • There is still no right answer to this challenge. When it comes to giving priority and value to any given channel, chances are that customers will be impacted by several impressions from your campaigns. To this day, there’s no right answer to the question. Data-driven models are likely the closest we’ve gotten to the solution, but even these have their limitations.
  • Context matters
    • Your vertical will have a certain audience profile. Understanding what media they consume, in what mindset, at what time of day (and more) is vital. TV might be your best bet, or it might not. Chances are that making broad assumptions with one variable only will have a high error margin, so context needs to be present in the analysis. 

Know what questions to ask 

A well-thought-out media strategy will be present in a wide range of ad inventory, and attributing value to each is a difficult task. So, here’s a list of questions that you should answer candidly before we start:

  1. How does your business perform without advertising?
  2. What role does paid media play in your marketing strategy?
  3. Do you have a single source of truth for the target key performance indicators (KPIs)?
  4. Has your strategy been consistent across different geographies and demographics? 
  5. Do you have strong and accurate forecasting capabilities for your business performance?

Build a reliable picture 

Understanding what you should expect when you don’t invest in advertising allows you to establish a very important baseline of expected business performance. A single source of truth for key business metrics provides an unbiased and reliable data set of the desired outcome. 

Consistent strategy across determined variables gives you the ability to draw comparisons (essentially, fixing the strategy variable), and your forecasting capabilities are key to understanding when changes have an impact.

Below are some recommendations to gather the answers above and start putting together a measurement plan that gives you not only a more complete picture but actionable insights.

Set the right goals 

The starting point is to go through KPIs and ensure these are set to be aligned with the ultimate campaign goal. Some good examples of this include: 

  • First sale ROI vs. expected lifetime value
  • Cost of sale vs. cost of incremental sale
  • Reach vs. incremental brand lift 
  • Leads vs. final sales value

What all of these have in common is that, for a lot of marketing teams, the focus is on the former when they’re trying to achieve the latter. 

Start by defining your KPIs according to business objectives. Then you can start thinking about measurement plans.

Define your method

Remember the baseline? Time to put it to use. There are different ways to approach this, depending on your media plans. As Alfred Korzybski famously said, “The map is not the territory.”

The one we would recommend is geography-based splits — the process of finding different geographical segments that have a correlatable KPI behavior. The key is that KPIs are correlated, not equal, as they have a similar baseline with historical data (sales over time, investment over time).

There are, of course, other ways to do this. The most used in digital would be exposed vs. non-exposed users within a determined target audience. There are a couple of issues with this method.

Mainly, if your business does any offline activity, that won’t be accounted for or controlled as it was envisioned. If your business operates in physical stores, chances are that you won’t account for that in the measurement. Finally, it makes a very big assumption that audiences are homogeneous — spoiler alert, they’re not. 

Test, validate, repeat

That’s not to say that geography splits are flawless. They can’t account for human mobility between borders, and they work out of historical data to establish baselines. (If the financial markets teach us anything, it’s that past performance does not equate to future results.) And they require the inconvenient aspect of differentiating strategies between target locations.

But testing is a fundamental part of measurement methodology — so it must be done. You must test different media and mix up campaign messaging to create a sturdy, insight-based hypothesis. Only then can you validate your results.

But the testing must not stop there: A data-driven media budget must be driven by validated hypotheses. 

Whichever way you choose, the important part is that you do choose to act. Understanding the value of a channel or a campaign can help you sketch investment and return curves that will help maximize cross-channel buying. 

Joaquim Salgueiro is partner and media director at LINK Agency.

DataDecisionMakers

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even consider contributing an article of your own!

Read More From DataDecisionMakers



Source link

Spotify’s AI DJ Has No Soul Previous post Spotify’s AI DJ Has No Soul
Microsoft accidentally offers Windows 11 upgrades to unsupported PCs again Next post Microsoft accidentally offers Windows 11 upgrades to unsupported PCs again