For whatever reason, marketers, particularly digital marketers, hate projections. When it comes to literally anything else, data is our bread and butter. So what is it about future performance that causes us such unease?
Past performance is set in stone – it tells us cold, hard facts. From past performance data we can easily (or not so easily) determine next steps and create a plan of action. However, there is a lot of uncertainty with projections. There is little way to know if this month or season will perform just as last month or last year did – just as there was no way of predicting the global pandemic that sent your accounts into freefall.
Additionally, there is a lot riding on projections. If a client or colleague is asking for performance projections, it is usually to determine investment levels going forward. Perhaps upper executives are discussing increasing a budget, or they are trying to gauge the performance of a new marketing strategy. Whatever the reason may be, accuracy is important.
So how do we create projections that are accurate and we feel confident in presenting to decision makers? Here are the top guidelines to follow when putting together projections:
1. Use past performance
If you have previous data available and are simply generating future numbers for increases or decreases in spend based on current campaigns/settings, this should be pretty easy. To use this data most accurately, you’ll want to look at both your last 30 days data, as well as quarterly data from the previous couple of years.
As traffic and conversion volume is directly contingent on spend levels, which may change year-over-year, you will want to look at rates instead – CTR, CPC, CPA. Take a look at your past 30 days data, find these rates, and enter them into a spreadsheet. Using these rates find the volume for impressions, click, and conversions based on your desired spend levels using the following formulas:
- Conversions = Cost/CPA
- Clicks = Cost/CPC
- Impressions = Clicks/CTR
Now you have the base of your projections for the upcoming timeframe. Next, use the previous years’ data to look for seasonal patterns in performance. Has there been a consistent, yearly change in performance during the time period you are projecting for? If so, you’ll want to factor that average change into each of the above metrics and rerun the numbers. If not, you are all set!
2. Use industry benchmarks
If you are making projections for a new client, or for a new campaign, without any previous data to work with, industry averages are perfectly fine to base estimations off of! This is particularly helpful with Display campaigns, where there aren’t defined keywords to bid on. Industry averages are easy to find across the internet and tend to be updated regularly.
*Pro tip: Even if you have previous data, taking a peep at industry averages is a great way to back up your projections.
3. Leverage Google tools when possible
Another option to investigate potential performance of new campaigns is using Google’s keyword planner. This is about as close to the source of truth as you can get. If you have your keywords for the new campaigns built out, you can simply plug them into Google and get the expected CPC, CPA, CTRs, CVRs, and even volume.
4. Don’t be afraid to adjust projections
When putting together your projections, keep in mind that they are just that, projections. While you’ve grounded your projections in data, you are perfectly justified in adjusting them based on your expert knowledge and past experience. For example, if you are projecting potential results of utilizing a new bidding method or new strategy, it may be impossible to quantify that exact impact it will have on your campaigns. While you may not be able to say “restructuring our account will increase conversions by exactly 23.6%,” you can say that based on your previous experience you’ve been able to increase conversions by approximately 25%. Whenever you are doing this, try to lean on the conservative side. No one will be upset if you exceed your original estimates!
5. Be able to explain your data
Following the guidelines I just layed out will help ensure accurate projections. However, being rooted in data, unfortunately, that doesn’t always make them right. That is okay. Decision makers know that projections are estimates, not guarantees. There is no fault in providing data that ends up being wrong, so long as the reasoning and logic behind it is solid.
In order to confidently present your data, make sure you know your metrics and be able to explain where your numbers came from. If you had to adjust your numbers or rates based on intuition rather than hard numbers, say so! If you had to use industry averages rather than account specific performance, say so! If you are upfront about your numbers and method, executives will expect a margin of error, rather than be surprised by it.
Long story, short – don’t let projections intimidate you. Pull hard numbers when you can and trust your expertise when you can’t. Be honest and set expectations accordingly. As long as you can defend your numbers, you really won’t have to. Good luck!