How to Plan Your Advertising Budget and Calculate ROI and ROAS

“How much should I be spending on digital ads?”

It’s one of the top 5, maybe even top 3 most popular questions I get from prospective clients. The interesting thing is that usually the person asking the question wants a fixed monthly number to enter into a budgeting spreadsheet.

Prior to doing a more in depth analysis, the best answer I can give is, “As much as possible, assuming you have a path to being profitable.”

But that’s not my complete answer. 

Start with a Fluid Mindset and Flexible Budget

When it comes to budgeting for digital advertising, your mindset needs to be fluid. A fixed annual marketing budget mindset is no way to manage digital advertising. In other words, you should think of your digital advertising budget as a range of possible spending that (if you are doing it correctly) can change based on holidays, seasonality, day of week, time of day, consumer behavioral patterns, and even changes in weather.

You’ll start with small doses of testing. As you test and collect data, you’ll see trends and behaviors that can be leveraged. Once you unlock a reasonable CPA or CAC (reasonable as determined by your LTV), you can focus on conversion rate. Once your conversion rate is in a good spot, you can focus back out on the full picture of your ad spend with ROAS. For every dollar you spend on digital ads, how much is coming back in sales revenue? Then you can zoom out further to look at your ROI. Once you crack profitability on your digital ad campaigns, you can ratchet up spend incrementally to 2x, 4x, or 10x what you were spending in month 1.

Confused by those acronyms? Don’t worry — we’ll define those terms and explain the process, so by the time you’re finished reading, the above paragraph will make perfect sense.

Here’s what we’ll be covering in this article:

• How your business goals affect your ad strategy and spend
• How testing helps you unlock the most profitable and efficient ads
• How to validate your efforts with metrics and data
• How to scale up your efforts and stay on top of the game

Determine Your Business Goals

Before even thinking about budget, the first step is to define your business goals.

Are you willing to grow at the expense of profit? The answer to this question determines how you move forward and your overall strategy. If you are willing, you can operate on thinner margins for a while and focus on acquiring and retaining customers. If not, you’ll want to start with a more conservative advertising strategy, focused on efficient campaigns.

The second thing to decide is what are your marketing goals? What are you trying to achieve with your advertising strategy? Is it brand awareness, website traffic, leads, or purchases? Your objective will determine what your advertising strategy looks like and the size of your budget.

Start Small and Test

Why Do I Need to Test?

Testing is important to determine which ad sets are most profitable, so you don’t waste money on inefficient or ineffective ads.

What Am I Testing?

A lot goes into testing ads, as there are many variables to test. Audiences, targeting, keywords, messaging and positioning, ad copy, the call to action, visual design and assets, landing pages,  and many more factors can be tested to ensure the campaign is as efficient and effective as possible. Of course, tests should be run scientifically, with only one variable per test and the rest of the ad as a control.

What Does a Legitimate Test Look Like?

A legitimate test is one where you can see a clear winner and can trust the results of the test. This means it’s statistically significant. Statistical significance is the likelihood that the results from your test will continue after the test — in other words, it’s the likelihood that your results are accurate. A good benchmark to shoot for is a 95% confidence rate that your results are accurate.

Minimal data requirements are another important factor. Online behavior isn’t the same from day to day, or even within the same day. For example, people have different habits on a Sunday versus a Wednesday. So tests need to run long enough in order to gather enough data to eliminate variety in behavior.

The main factor in a legitimate, statistically significant test is volume. If you have a relatively large amount of ad spend, impressions, and conversions, your tests will be able to show clear results within a short time frame. If you have a small amount of traffic and conversions, or a long lead time, it will take longer for tests to show clear results.

When testing ads, it’s best to measure solid, bottom-of-funnel metrics like conversions and purchases over top-of-funnel metrics like impressions and clicks. The reason being that you’re tracking the funnel all the way from ad impression to conversion, as opposed to tracking half of the funnel, which leaves you with an incomplete picture of how the ad is contributing to your bottom line. However, with small campaigns that have few conversions, you can look at those softer metrics to get a sense of how ads are performing without having to wait months to get significant conversion results.

Once a test is done, you should have a clear picture of which variable was most effective. Then, you can extrapolate that variable to other ad sets and build on it to continue testing and optimizing other aspects of the ad.

How Many Channels to Test?

It’s a good idea to start by testing multiple channels in order to see which platforms are most effective at reaching and converting your target audience. Each channel has a different audience — think of the difference in demographics between Facebook users and Snapchat users. Channels are also differentiated by user behavior and/or intent. For example, someone searching for a product on Google is likely in a different stage of the buying process than someone seeing an ad pop up on a social network.

Typically, we’ll start with at least Google and Facebook. People searching for products are already primed to buy them. We love to use search as a platform to test because of the volume and high user intent. It’s useful to test product positioning and messaging, and use the knowledge gained from test results to inform the brand’s web presence, and the marketing and advertising strategies. Facebook is another great platform for getting early insights because of the flexibility and power of their audiences and targeting.

The beginning of any ad campaign is all about learning. Testing multiple channels at once accelerates the learning and results in lessons that can continue to be used throughout the lifetime of the campaign.

Once you understand how the different channels reach different audiences, stages of the buying process, and convert differently, you can start to apply those lessons by adapting campaigns, eliminating wasteful channels, and experimenting with new channels — for example, if Snapchat is working well, then Tiktok (which appeals to a similar demographic) might be a good channel to test next.

Validate With Metrics

The key to understanding how your digital advertising strategy is affecting your bottom line is to look at the metrics. These will help you trace the buying process of your potential customers all the way from initial contact to purchase to retention, measure the effectiveness of your ad campaigns, and optimize your campaigns for peak effectiveness.

Marketers tend to throw around a lot of acronyms, so let’s take a minute to define some of the common terms here.

Should I Use ROI or ROAS?

This is a common question that we get, and the answer is, there’s a time and a place for both.

ROI takes all expenses into account, so it’s great at looking at the big picture, and understanding how your investments lead to your bottom line.

ROI = Net Profit / Expenses

Read more about what goes into calculating your ROI in our article about how to predict ROI with confidence.

ROAS is more granular, and is obviously looking specifically at ad spend. It’s calculated by looking at revenue over ad costs. ROAS is useful for gaining a broad understanding of the efficiency of your ad spend, and especially for knowing which ads are performing best.

ROAS = Revenue / Ad Spend

The issue with ROAS is that to get an accurate picture, you need true attribution set up. True attribution means that you can track customers all the way through the funnel, from clicking on an ad to making a purchase. It’s calculated by looking at the revenue generated by your ad spend, so obviously, you need to be able to tell how much revenue is generated by ad spend. Many companies don’t have proper attribution set up, so it can be hard to properly calculate ROAS.

What Should My ROAS Be?

The answer is, of course, that it depends. Because ROAS doesn’t take expenses other than pure ad spend into account, every company’s ideal ROAS will be different, depending on their operating costs and marketing expenses. At the end of the day, you want your ROAS to be positive — that is, you want to be making more money than you’re spending — and you want it to be as high as possible, to maximize profits. The industry benchmark is between 3-4 ROAS, meaning for every dollar spent on ads, your revenue is between $3-4.

How Do I Know If My Strategy Is Profitable?

In addition to looking at ROAS and ROI, it’s good to look at additional metrics to evaluate the profitability of your advertising campaigns and overall strategy. Ultimately, how much money you spend on advertising comes down to your LTV and CAC. Just like ROAS (and everything else in the business world), at the end of the day, you want to be making more money than you’re spending. In other words, the cost to acquire a customer should be a fraction of your customer lifetime value. A good rule of thumb is that if your LTV is more than 4-5x your CAC, then you are probably underspending. It’s time to look at upping your budget!

The great thing about these metrics is that once you have your model nailed down, increasing growth or revenue is simply about pulling levers. To increase revenue, you can lower the CAC or raise some levers in the LTV, like average order value or number of transactions in the lifetime.

Here is an example of an ROI projection for one of our e-commerce clients. You can see the different factors that play into the calculation, from the fixed costs, to revenue per unit, to adjustable costs like CAC.

Below is projected ad spend for the same client. The cost per click, site conversion rate, and CAC are all subject to change, which is a good thing. Part of a marketing team’s job is to improve those rates: lower the cost per click and CAC and raise the conversion rate. As those move in the right direction, the ratio of spend to revenue will improve and each dollar you spend on advertising will go further.

Scale and Diversify

How Do I Know When to Increase or Decrease Spend?

With digital advertising spend, there’s always a sweet spot. Spend too much and there’s a tipping point of diminishing returns —there’s only so many people in your audience, and if they are seeing the same ads over and over, this will cause ad rot. Fewer people will click through, which will in turn hurt your quality score, resulting in lost impression share — in other words, fewer people will see your ads. If you spend too little, your campaign won’t max out its effectiveness. The sweet spot is different for every brand, conversion goal, and campaign, so it’s a matter of optimizing and tinkering until you find that sweet spot.

There are some general rules to follow, though: It’s good to increase spend when a campaign is optimized and profitable, as determined by your metrics of CPA and ROI or ROAS. It’s good to decrease spend if it’s being wasted on badly optimized campaigns — in this case, it’s good to decrease spend a bit while tweaking and testing to get the campaign back on track, then spend can be ramped up. This can also apply to badly performing retargeting campaigns: often that’s an indication that you need more brand awareness before you put more money into retargeting.

A very important point to keep in mind is that Facebook recommends making changes only in 20% increments: any shift larger than that will reset their algorithm, eliminating everything that Facebook’s algorithm learned about your campaign, and essentially resetting the entire campaign.

A Lesson on Diversification

As a final note, it’s important to not put all your eggs in one basket. The internet is ever-changing and if you’re overinvested in one channel, just one move by Facebook, Google, or a competitor could put you out of business.

For example: say your search campaign is performing really well. It’s efficient and well-optimized, with a low CPA and a high ROI. It’s bringing in tons of customers and fueling your business growth. What happens when a competitor starts bidding on the same keywords and drives up the price? If your strategy isn’t diversified across multiple channels, it could throw a major wrench in your profits.

Just like your investment portfolio, your digital advertising efforts need to be diversified to protect against unknown future developments or changes by ad platforms. For more reading on diversification, refer to our past article on Optimization vs Diversification.
So: You understand ROI and ROAS. You know your business goals and LTV (or you don’t know your LTV, and you’d like to). You’re ready to start spending on digital advertising. We can help! Contact us and we’ll set up a time to talk about how we can grow your business.