Lorem ipsum dolor sit amet, consectetur test 2 adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.

Picture of Jon Lorenzini
Jon Lorenzini

What Drove That Sales Spike: The Season, the Promotion, or the Ad?

Two ways to tease out the impact of advertising versus seasonality or promotions.

When you’re evaluating the true impact of advertising, you need to be sure you measure and model the impact of your ads, and the ads alone. To see how this could be a complicated feat to pull off, look no further than two of the biggest non-advertising factors in conversions: seasonality and promotions. For instance, if your sales spike this June, should your ad campaign take the credit – or was it primarily Fathers’ Day Shopping and your Fathers’ Day sale that brought shoppers to your store?  

Keep in mind that the impact of these factors – which, of course, often come together – can be huge. For instance, per Professor Dominique Hanssens, LiftLab Advisor and Distinguished Research Professor of Marketing at the UCLA Anderson Graduate School of Management, changes in price are 20 times more impactful than changes in advertising. That’s an enormous amount of “background noise” to work through when you’re evaluating the data.

Given the extent to which seasonality and promotions can drive sales, marketers have a clear interest in separating out the different factors at play to better understand true impact. To better decouple ad impact from pricing and promotions impact in your own analyses, start with the two strategies below.  

1. Flag Promotional (or Non-Promotional) Days 

 It’s challenging to distinguish the impact of ads from the impact of promotions running concurrently. But since your brand is in charge of the promotion calendar, it’s easy to know exactly when promotions run – and so it’s relatively simple to exclude promotion days from datasets entirely.  

If you run promotions relatively infrequently, note the day(s) you run a promotion and flag that time period’s conversion activity as not representative of the norm. With that flag as your guide, simply don’t incorporate those dates into your ad effectiveness models. 

Conversely, if you run promotions daily, as is the case with many of LiftLab’s clients, we recommend a kind of “negative” flagging, denoting the days you don’t promote.  

Whether you’re flagging days that promotions do or do not run, the concept is the same. Identify what the promotions norm is, note the days that deviate from the norm – and exclude those timeframes from model inputs. 

2. Go Cross-Seasonal for Experiments 

If you’re experimenting by pushing ad spend higher or lower, one of the least helpful things you could inadvertently do is push ad spend higher during a high-demand season or lower during low-demand times. Looking only at the conversions during those timeframes, you could easily be convinced—incorrectly—that pushing spend up causes sales to skyrocket, dropping spend causes sales to plummet, and that your advertising overall is far more effective than it actually is. The converse is equally true: If you drop ad spend in a peak season, you’re still likely to bring in organic sales, which could give you a false indication of the effectiveness, or lack thereof, of your advertising. To minimize these effects of seasonality, be sure to run tests and experiments throughout seasonal peaks and troughs, not just at one point in time. The more diversity you can get in terms of testing timeframes, the more closely you’ll be able to track the true impact of advertising. 

The Complications Get More Complicated 

Above, I’ve presented some universal approaches that, implemented correctly, can help you better decouple promotions and seasonality impact from the impact of your advertising program alone. To go further in teasing out the impacts, however, you’ll want to account for the particular nuances of your own business.  

For instance, you might be serving promotions on most but not all days, or you might be running constant promotions but vary the promotion size quite a bit from day to day. Additionally, seasonality in your vertical likely includes a mix of major and “minor” impacts—not just the huge influx of Fathers’ Day shoppers, but, say, a small but significant uptick in purchases around The Oscars or the Kentucky Derby. Any of these variations could impact the optimal strategy for accounting for seasonality and promotions in your marketing analysis. 

Practically speaking, this means you’ll want to see the above recommendations as a strong starting point for thinking through these decoupling problems – but it is hardly the last word on the matter. 

Working at LiftLab, I have the benefit of our advisory board of experts in the field, including noted academics who can help me work through these kinds of subtle decoupling problems.  

Since variations of these problems frequently appear across our clients, we’ve worked many capabilities into the LiftLab Marketing Effectiveness System. If you’d like to learn more about how LiftLab could help you isolate the impact of advertising from other external factors like seasonality and promotions, and more, book a conversation with someone from our team. It will likely be a useful conversation, whatever season you’re in.