How To Improve Ad Performance With Forecasting

Knowing tomorrow's cost per conversion and conversion volume feels like a cheat code to better performance, but how do you truly use it to drive ad account optimizations. In this post we'll breakdown what type of insights you can gain from forecasts, such as creative fatigue, seasonality patterns, and more, as well as what to look for to identify them. We'll also go over what is required from advertisers in order to get the most from forecasting.

Forecasting Requirements

When it comes to forecasting, your forecasts are only as good as your data is. In order to effectively and accurately generate a 7 day forecast, Winch requires a minimum of 12 months of ad spend data. Forecasts can still be generated with less data, but as you decrease your sample size, you also decrease forecast accuracy. For best possible results, forecasting is most suitable with ad accounts who have a stable account structure, such as utilizing evergeen campaigns.

The Types Of Insights Forecasting Can Drive

Creative Fatigue

Creative fatigue is something every single advertiser on Meta can, and usually will experience. Easily identified by a slow and steady decline in click-through rate's (CTRs) over an extended period of time, with rising cost per conversion, and possibly compounded with rising cost-per impressions (CPMs) as well. Without forecasting, it usually takes marketers a couple days to first identify it as a possible issue, and another couple days watching performance decline to confirm it is indeed creative fatigue and not just a small hiccup in an otherwise well performing campaign. By identifying say a consistent rise in cost per conversion in the forecast, you're able to remove the hesitation from the equation and act quicker on rolling out new ads. Not something most small businesses spending $100 a day should stress over, but spending thousands per day on underperforming creative is enough to make forecasting drive real ROI.

Seasonality

Similar to creative fatigue, by forecasting cost per conversion, you're able to identify seasonal patterns within the days, weeks, and months of the year you can use to your advantage. For example, if your forecast is showing a large spike in cost per conversion over a holiday, you can use that insight it to momentarily pull back on ad spend. Or if you're trying to compete during cyber monday, you have a better idea of predicted conversion volume based on previous years patterns, and the patterns leading up to this years holiday all being accounted for. Forecasting and harnessing seasonal patterns are things advertisers at any spend level or industry could benefit from.

Proactive Optimizations

Marketers have almost all relied solely on historical data to drive ad account optimizations and decision making. Simply having a forecast is an added layer of insight into the dozens, if not hundreds, of decisions we make in our ad accounts every day.