Table of Contents
Today we’re walking into a minefield: brand search. But instead of offering a one-size-fits-all cursory analysis, we’re going to answer the question, “Should I pay to advertise on my brand terms?”
Brand search is a hot topic in paid media, and people have strong opinions about it. Understandably so, it is a deadly mix of economics and egos that calls into question the efficacy of the work many of us do. But whether you believe that brand advertising has no measurable benefits, or you’re a brand search evangelist, I challenge you not to paint brand search with a broad brush.
Further down this article, we created a tool for measuring the correlation between brand spend and return. Jump straight there, or enjoy the build-up.
Your Brand, Your Bottom Line
Each of our clients exists in their own competitive environment. The impacts of brand search should therefore be weighed individually. While many lists that outline the benefits of brand search exist, these marks in the “pro” column are not universal. They will not apply in all circumstances. And each of them has a counterpoint in the “con” column.
Take, for example, possible benefits like:
- Brand search lets you avoid losing clicks to your competitors
- It’s cheap, and the return on investment is high
- Ads increase the total amount of clicks, so brand spend will help drive more traffic
In the first case, it follows that if you don’t have competitors on your terms, you won’t lose traffic and revenue to them. In the second, if you consider that brand search may be cannibalizing organic traffic, then your return on investment is lower than it would be if you didn’t pay for those users. And finally, while studies from the early 2010s show a general uptick in total clicks when a company has both an organic and paid listing, I haven’t seen any updates to this data across industries and countries for the intervening decade. Take it with a grain of salt.
All that is to point out the other side of the coin, and to say that each company’s circumstances are unique. You may see incremental return from brand search if your large competitors are absolutely devouring your search results, and it might not be worth it if you’re a small, local business with relatively few competitors.
The hard part is knowing what’s true for your business.
Brand Spend Correlation Test
To really know what brand spend is doing for you, you’ll likely need to run a full test -- you’ll likely need to pause your brand terms for a time. It’s a risk, but understanding the true impact of brand spend requires making some changes to our “X” variable. However, if you’d prefer to explore the data first, you can start with some correlation tests.
Correlation measures how changes in one variable affect changes in another. In this case, we want to know whether changes in brand spend affect changes in our total revenue (or leads, or sales, or whatever your key metric is).
To do that, we’ll run what’s called a Pearson correlation test. This measures the strength and direction of a relationship between two numerical variables: age and salary, shoe size and height, etc… The output is a value between -1 and 1, where a value closer to 1 represents a strong correlation, and the sign represents either a negative correlation or a positive correlation.
We also want to know if our relationship is significant (if it is due to something other than chance). And we want to see just how much variation the relationship explains. These numbers are a bit more complicated, but the tool below offers more context and explanations in the results.
Those numbers give us information about the strength of the relationship between brand spend and total revenue. They tell us whether the relationship is due to something other than chance, and they show us whether changes in brand spend explain large portions of changes in total revenue.
Uproer’s Brand Correlation Tool
I wanted to make a full-scale tool that churns out the report right here on this page, but that quickly went above my head. In any case, we’ve created a quick-and-dirty spreadsheet that will run through the Pearson correlation test for you. All you need is the following:
- Your historical brand spend
- Your total revenue over the same period of time.
For additional context, you can do the same process with this tool. Just make sure your data does not have commas.
Here's a preview of the types of results you'll see when you bring your data into this spreadsheet. Check out the "How to Use This Spreadsheet" tab for information on how to interpret your results.
Things To Consider
This test does not offer a definitive answer. To truly understand the impact of your brand spend, you should run full-scale experiments, with control groups and test groups over the same time period. That will equip you with cleaner data you can use in more nuanced statistical analyses.
However, even with those analyses, there is still room for errors. Some common errors stem from:
- Data Quality & Quantity Issues
- You need to make sure you have enough data points to reach an actual conclusion. And you need to make sure that your data is clean enough to work with.
- Measurement & Attribution Issues
- If you don’t have confidence in your measurement, then you could be misrepresenting the strength of the relationship at hand. This is especially pertinent when looking at conversion or sales data from online analytics, because attribution models differ across channels and can lead to double-counting that muddles our data.
- Brand spend and total revenue could both be correlated with another variable, which we do not account for. This could make it appear that our variables are related, when they are not.
- Cum Hoc, Ergo Propter Hoc
- This is a logical fallacy that means “with this, therefore because of this.” For example, brand terms may spend more if your brand gets more recognition. If you measure this relationship over a period in which your brand was growing in popularity, you will likely see a positive correlation even though that growth was not necessarily the result of brand search.
- Correlation Is Not Causation
- Here’s the big one. We are measuring whether your total revenue and total brand search spend are correlated. That does not tell us for sure that there is a causal relationship between the two.
Test & Analyze Brand Performance
Like I said; it’s a minefield. And because it’s a minefield, we should avoid sweeping generalizations about brand spend. Instead, each company should deliberately evaluate the incrementality of brand spend for themself.
Yes, it’s harder to find consistently good non-brand traffic. Yes, your return numbers will drop. But that’s the wrong way to look at brand spend. If brand spend isn’t helping you reach your broader goals, your return numbers are fluff. Cutting brand search spend frees up time, energy, and budget to focus on more impactful paid search efforts -- efforts that can drive truly new customers.
So let me answer the question head on: “Should I pay to advertise on my brand terms?”
Yes, but not indefinitely. Do it for a period of time to measure the impact it has on total sales, revenue, etc… Then take a step back to analyze that data. If it’s not increasing the numbers that matter for you, then pause it and allocate that budget somewhere else.
Cue Minesweeper theme music.