Podcast

Predictive analytics’ power to forecast your marketing future

Data forecasting – or predictive analytics – can be a powerful marketing tool. In essence, it can give an idea of what is realistic, based on data, regardless of anyone’s own perceptions or “gut feelings.” It’s a prediction of what is going to happen, even in some cases, that may not paint a very good story.

However – done well, predictive analytics can position a company or a media plan to accurately interpret financial goals based on data instead of just intuition alone. In addition, this approach can further contextualize any new variables in play alongside more traditional approaches or datapoints.

In this episode, we discuss how effective implementation of data forecasting as part of your analytics and reporting can establish benchmarks that empower a company to make category-leader-caliber decisions based on sales and data,  and not just reacting to what competitors are doing or CPG trends.

Listen here (Subscribe on iTunesStitcherGoogle PlayGoogle PodcastsPocket Casts or your favorite podcast service. You can also ask Alexa or Siri to “play the Uncovering Aha! podcast.”):

Welcome to Callahan’s Uncovering Aha! podcast. We talk about a range of topics for marketing decision-makers, with a special focus on how to uncover insights in data to drive brand strategy and inspire creativity. Featuring Jan-Eric Anderson and Zack Pike.

Jan-Eric:
Hi. I’m Jan Eric Anderson, Head of Strategy at Callahan.

Zach Pike:
I’m Zach Pike, Head of Data at Callahan.

Jan-Eric:
Today on the podcast, Zach and I are going to talk about a topic that we love, which is forecasting.
Let’s start with defining what forecasting is just to get everybody, ourselves as well as anyone listening on common ground about what forecasting actually is.

Zach Pike:
Yeah. Forecasting is a subset of predictive analytics. The way that I think about forecasting is it’s my watermark on what I expect to happen, all else being equal. It’s my best estimation of whatever I’m forecasting. If it’s revenue, if it’s customers, if it’s email clicks, it’s my best estimate, all else being equal, of what that number’s going to be over whatever time horizon that I’m estimating it. Typically, it’s a weekly or monthly forecast, but it gives me a watermark to measure myself against.

Jan-Eric:
Okay, so forecasting is a subset of predictive.

It predicts the future. What’s going to happen if everything else is the same next year on whatever metric we’re looking at, let’s say we’re looking at sales, it gives us our best guess about what we’re going to have for sales. Why is that important? Why does that matter?

Zach Pike:
I need to know what to expect, right?

What often happens is we walk into the year just hoping that we’ll do better than last year, or we’re going to attain some target that was set. Building a legitimate forecast is an exercise at looking at history and saying, okay, based on what I’ve done in the past, and based on what I know about what’s going to happen next year and how that’s different from what happened in the past, what do I think I’m going to do.

I mean, that is a much better mechanism to benchmark yourself against than your comps from last year or than a target that was set by the board.

Jan-Eric:
In essence, it gives us an idea of what we think is realistic. Regardless of what anybody else says. Or what wants to have happen. This is what we predict is going to happen. Even in some cases, that may not paint a very good story.

Zach Pike:
Right, yeah. Sometimes it brings you to reality. If you think about the typical forecasting method of most companies, it’s not actually forecasting, it is taking the number and spreading it out across the year.

Zach Pike:
Right. It doesn’t happen like this, I’m going to dramatize it a little bit, but the board got in a room and decided we need to hit a hundred million dollars next year, and this year we did seventy million. So, that’s the forecast.

Then a finance guys take the hundred million a spreads it out with some logic to say, I can do ten million in January, and twenty million in February and it will be spread out this way across the stores, and the business units and things like that, and they call it a forecast.

Jan-Eric:
It’s not a forecast, it more of a requirement. To get to a goal.

Zach Pike:
I don’t say this from an ivory tower because I’ve done this, I’ve been that finance guy spreading this out across sales reps and business units and everything else.

But yeah, to your point, it’s not a forecast it’s a goal and a logic based method on how that goal could be attained.
When we talk about forecasting it’s an objective unbiased approach to saying what is really going to happen next year.

Zach Pike:
And oftentimes… I’ve never done it when that forecast actually lined up with the goal of the company.

Jan-Eric:
So back to your example, if we did seventy million last year and our goal is a hundred million this year, and the analyst, or the finance person has spread that incremental thirty million dollars of growth across all the months to give us a plan of how we need to get there.

If we have a forecast that shows that we’re actually going to be at $60 million in revenue this year, then we don’t need to grow 30, we actually need to go find 40. That paints a very different, that paints a very, very different picture.

Zach Pike:
Absolutely.

Jan-Eric:
That flat would actually be up.

Zach Pike:
Especially, and we do this… if you were in a marketing role, that’s very important information. In your scenario you just mentioned, what I would think was happening to the business is it was on a multi-year decline.

If I did seventy last year, I think I’m going to do sixty next year, that suggests that prior years I was doing more than seventy and I’m kind of on my way down.

That’s very important because that means, sales team, marketing team, product management team has to reverse the direction of the company. They have to reverse a trend which is very different then growing faster than you were already growing before. That might influence how you go about it and how much money you need to do it, who all needs to be involved and all that fun stuff.

Jan-Eric:
Understanding the difference between a forecast and a target starts to set the challenge up for a marketing team. What are we… what’s our gap that we’ve got to make up. There’s always growth that’s involved-we need to do new things, we need to be trying new things, we need to be testing new things, we need to be generating revenue that we didn’t have a year ago by… not by doing all the same stuff, but by doing new things.

I think that starts to tee up in the next benefit of forecasting is that it put you in a position to be able to measure the impact of tests-when you get into control and test scenarios, where you’re testing new variables or new marketing tactics to understand how they really work. You don’t look at it versus last year’s sales.

You look at its results compared to what you forecasted for this year.

Zach Pike:
Right. It makes you a lot smarter. What often happens in AB testing or control and test group segmentation or even multi-variant testing is you’re looking at your stimulus against the control and just hoping that it’s better. You’re measuring how much better and all that stuff.

At least in my experience it gets a lot more valuable when you do that, but you also have that context of what you expected to happen for your test group. Now, it’s okay I was better, but I was also better than what I expected this group to do anyways. You have the context of ‘I reversed their direction’ or ‘I just accelerated their growth’ or whatever.

We’ve talked about this a lot on the podcast. The goal of analytics is to improve confidence in decisions, the more little attributes you can insert into that confidence equation, the better off your results are.

Jan-Eric:
Yeah, it certainly puts in context the results too. I remember an old buddy of mine, a friend from college was in a role working for an analytics company, and they were doing predictive modeling and establishing forecasts for a big box retailer and their focus was on flat screen TVs.

It was fascinating to me. He was sharing with me how they looked at sales performance by different stores. They were at the corporate headquarters and their central head of sales, they were building incentive plans for their sales reps in these big box retailers in the different locations because they were incented to sell more flat screen TV’s.

There was a group in Chicago that was selling fifteen percent more TV’s then they had been the prior year and they were getting paid handsomely – good bonuses because they were performing so well.

Then there was a group in Iowa that was plus 5 percent on their TV sales and that was not enough to earn them incentive bonuses, they hadn’t crossed a threshold.

My friend’s analytics company came in and showed in hindsight, forecasting, and the guys in Chicago should have been twenty percent higher based on economic conditions in the area where they were and the guys in Iowa should have been 5 percent less.

It shined a completely different light. The big aha for the company was they were rewarding the wrong guys.

Zach Pike:
Yeah, happens all the time.

Jan-Eric:
The wrong sales people were being be bonused based on their performance because there was not a forecast upon which that they were basing their incentive plans – it’s really fascinating.

Zach Pike:
Think about if you were in a marketing leadership position and you had that context-you had the context of who, whatever you who is, if its business unit or area of the country or whatever, should be doing better and who should not be doing better.

Jan-Eric:
Who do you hold up as an example-at your national sales meeting-as this is the team that gets it, this is the team that we’re recognizing-and this is the team that we’re going to model our future training off of.

It’s not the guys in Chicago, no offense to the windy city. It’s the team in Iowa in that case, was the example of the team that was really performing well.

Zach Pike:
Yep. In sales meeting, but also the boardroom.

Jan-Eric:
Right.

Zach Pike:
You’re a CMO, you have all this pressure to perform better than you performed last year, or 2 years ago or whatever. What if you had the context to say what we have to believe to actually perform at the levels you’re expecting me to perform.

Jan-Eric:
Yeah.

Zach Pike:
It changes the conversation. It puts you in a more strategic, director position, which is unfortunately not a place that many marketing leadership sit today.

Jan-Eric:
It clears the smoke right, and it gives you a better read on what you’re really looking at.

Zach Pike:
Yep, yeah.

Jan-Eric:
We’ve talked about the… a benefit of forecasting here is that you can avoid that trap against the target and not misinterpreting what the financial goal is-versus what you think is actually going to happen. Avoiding that trap’s a big benefit.

Another big benefit was certainly being able to understand better context on new variables that you’re bringing in or new testing that you’re doing in marketing.

Jan-Eric:
The third piece is just having benchmarks that are specific to your business-rather than relying on what the category is doing, or who your biggest competitors are and assuming that you have to follow their same trends.

Zach Pike:
Yep. This just came up in a meeting I was at yesterday was the whole benchmark discussion again. It’s something that every marketing person deals with all the time is well how does that compare to benchmark. Whatever the metric is, how does it compare to benchmark and you can try to smash the benchmark argument with stuff we’ve talked about on the podcast in the past which is what does the industry mean, a benchmark is just an average, do you really want to be average all that stuff.

The strongest mechanism we’ve found is well, I could use a benchmark, but here’s what I thought was going to happen, and here’s my logic on how I got there. That seems a lot stronger than a benchmark.

Jan-Eric:
No question about it. In the conversations we had, I think our clients have found it very useful as well. It has certainly been useful for the work that we do.

Zach Pike:
Yeah, yep.

Jan-Eric:
To have to be grounded in a forecast but certainly seen the value from our clients as well.

Anything else you want to add before we wrap up?

Zach Pike:
No. The only thing I would say is to take your time with the forecast and be honest about it. If you’re going to actually do a forecast, then do it, and communicate it and don’t hide from it.

Be honest about it and if people don’t like it then they can not like it. If you can show them the logic you used to get there, or your analytics person used to get there, show them what the number is and even if they throw it in the trash, at least you’ve planted the seed to say what we think is going to happen, reality is different.

Jan-Eric:
You’re objective in forecasting is to be as accurate as possible, not to paint a rosy picture.

Zach Pike:
Exactly, right.

Jan-Eric:
You can caveat it that this is conservative or that this is aggressive-you can outline assumptions, that’s all fine but the objective of a forecast is to most accurately predict the future, not to tell somebody what they want to hear.

Zach Pike:
Exactly, yep.

Jan-Eric:
Yeah. That’s probably the most important note to leave on.

Zach Pike:
Yep.

Jan-Eric:
Zach, thanks for joining us it’s been great as always.

Zach Pike:
Yep.