The problem with paid media forecasting

We've all been there. The board has asked you to generate more sales.

You've been told you need to get 500 more sales with the extra ￡20k of budget. After all, you generated 1000 sales with ￡40k budget; what could go wrong?

You then go to your agency or look for a new agency to help you scale and tell them we need 500 more sales with the extra budget.

The agency delivers a forecast showing the extra sales for the budget.

A few months pass.

The forecast doesn't come off. You're annoyed at the agency for not delivering. The board's annoyed at you for not hitting the sales targets.

Where did it all go wrong?

This brings up three main issues I see happening frequently regarding forecasting.

Agencies need to stop creating forecasts to fit the brief

You see this happen often in existing client relationships and the agency pitch process.

An agency is told that we have x amount of budget and need x amount of sales; therefore, they create a forecast that shows that.

We need to focus on creating accurate projections and setting more realistic expectations for both the board and the CMO.

There needs to be an emphasis on pushing back and focusing on accuracy instead of fitting projections to the brief, irrespective of the likelihood of them occurring.

Agencies need to be better at educating brands about diminishing returns.

On the agency side, I can speak from experience. It's demotivating to be given unrealistic targets, especially when you have yet to be consulted, which, in my experience, is often the root cause.

It's understandable that senior stakeholders may not clearly understand the diminishing returns you get from paid media advertising.?

There is also a tendency for C-level board members to be sucked into the planning fallacy. Alas, it's a trap. One I've fallen into in the past. Daniel Kahneman sums this up succinctly in his book Thinking,?Fast and Slow.

"When forecasting the outcomes of risky projects, executives too easily fall victim to the planning fallacy. In its grip, they make decisions based on optimism rather than on a rational weighting of gains, losses, and probabilities. They overestimate benefits and underestimate costs. They spin scenarios of success while overlooking the potential for mistakes and miscalculations."

Practitioners are acutely aware that the incremental cost to acquire more customers increases significantly as you spend more.

It may seem reasonable for a senior stakeholder to say, we've got 1000 sales this month for a ￡25 CPA.?

Why can't we get 500 more?

Therefore, I think it's vital for an excellent agency-brand relationship for agencies to be proactive in providing projections at different spending levels so that boards and senior marketers can make informed decisions.

Brands need to get better at understanding and challenging the methodology behind agencies' forecasting.

There is definitely a lot that agencies need to work on, but forecasting takes two to tango. Brands need to interrogate the forecasts agencies present better and better understand the methodologies used.

I can't recall a time when a brand's senior marketing manager asked me how we arrived at the forecast. Some better marketers ask a few pointed questions and interrogate a few stats if they don't think they align for one reason or another.

This is particularly pertinent when it comes to pitching and the forecasts provided. There is undoubtedly a danger that agencies may look to over-forecast to win the pitch. Of course, you want to put your best foot forward as an agency, and the bigger your projections, the more likely you are to get hired.?

However, you also need to ensure you can deliver them. I think, in many cases, agencies get a little carried away with these.

You certainly want to ask some tough questions about how they arrived at the figures.?

You'd expect the agency to look at YoY data to arrive at the projections. If they're looking at scaling spend, they should use tools like a performance planner to provide CPC estimates at different spend levels.

Suppose agencies expect to see improvements in certain metrics like CTR from testing ad copy, for example. You'd want to ensure that they've seen similar results with existing clients in a similar sector.

The best questions to ask about the forecasts are about the uncertainties. Can the agency explain the uncertainties in the models that they've used?

It can also be helpful to ask your agency to provide a couple of different forecasts, factoring in a more optimistic view where things perform a little better than expected and a second, more pessimistic forecast looking at reasonable worst-case scenarios to set the board's expectations.

If you want to learn more about forecasting, you can watch our recent webinar here and download our paid media forecasting template here.