Predicting ROAS and Launching Better PPC Campaigns

ROASReturn on Ad-Spend, or ROAS, is a metric calculated by dividing the revenue generated from an ad campaign, by the cost of that campaign. It can be applied to any sales situation that has an advertising spend, even a bake sale:

Lets say your mother gave you a pan of brownies to sell at the bake sale, but it costs $50 to get your booth and marketing materials set (we’ll call $50 your ad-spend).

If your brownies manage to produce a total revenue of $50, you’ve broken even, or gotten a full return on your ad-spend.

If you had made the brownies yourself, and they’d cost you $50 to make (we’re talking primo brownies here), then your total investment is actually $100, but ROAS only takes into account ad-spend. So in the same situation you’re still getting a full return on your ad-spend, just not a full return on your investment.

A handy exercise to go through with new clients before you launch a PPC campaign is to calculate the potential ROAS of the keywords you’re targeting.

There are a few unknown, and a few fuzzy variables – we’ll have some tools to help us come up with potential traffic and potential costs, but we’ll have to use some educated guessing to come up with reasonable estimated ranges of conversion rates and click through rates – in truth, the conversion potential of your keywords, through your ads, to your landing pages, for your clients products, are up to you and your marketing team. Sorry ’bout that.

Calculating ROAS:

In order to calculate a predicted ROAS in PPC advertising you need to know the revenue that each conversion brings for each product you’re marketing (which could be associated with the keyword, or ad-group level), the potential cost of the campaign, and a half-decent guess at what kind of a conversion rate you might achieve. Once you know what a conversion is worth, all of the info that you need is available in rough form from Yahoo and Google’s keyword tools.

Yahoo give us the estimated number of clicks and cost per click (CPC) for a keyword so we can calculate the expected cost of the campaign per month. Adwords give us approximate search volume along with the estimated CPC, which we can use to determine a rough estimate of traffic at different click-through-rates, and what it will cost us. If we supply a hypothetical conversion rate to calculate conversions, and we know the company’s revenue from each conversion, we have what we need to calculate ROAS. So like I said, it gets fuzzy, but it’s still half-decent for predicted information.

These simple equations are expressed like this:

To further the confusion of the ROAS metric, it is commonly calculated in two *different* ways. One way subtracts the PPC ad-spend from the revenue right in the calculation, the other way does not. Let’s look at each equation and how to interpret the results.

Method one:

With method one revenue of $100 and an ad-spend of $100 would produce a full return on ad spend of %100. So with this method of calculation seeing 100% ROAS simply means you’re breaking even. A 50% ROAS would mean you’re only recouping half of the ad-spend you’ve put out. With this equation in a situation where you have $200 revenue and $100 ad-spend, doubling your ad-spend is expressed as an ROAS of 200%.

Method two:

In this formula the cost is subtracted from the revenue, forming a profit metric that is relative to the ad campaign. Plugging the same numbers into this, $100 ad-spend with $100 cost, break-even comes out to 0% ROAS. With this equation, in a situation with $200 revenue and $100 ad-spend, doubling your ad-spend is expressed as and ROAS of 100%.

Be sure you know which type of ROAS you’re looking at, because obviously the interpretation is different for each.

Experiment with a Spreadsheet:

I’ve put these equations into a simple spreadsheet (download). Here’s what it looks like:

The heading ‘revenue of campaign’ might refer to a whole PPC campaign, an ad-group, or just a keyword – it is the revenue per conversion, times the number of conversions (and depending on your preference, minus the cost). You may want to set the sheet up based around just one variable, such as conversion rate:

Or cost per click:

Useful, but comes with caveats:

The basic approach of pre-checking keywords for ROAS can help you launch campaigns that stand a better chance of being profitable quickly, rather than analyzing data after some time (and some money!) has passed. It can let you know going in, all else equal, approximately what costs-per-click should stand a good chance of performing well. The problem is, all else is never equal.

A number of variables affect your ROAS, and tweaking CPC is only one part of a comprehensive management approach. Some argue that ROAS should only be used as a general indicator because it is too broad in scope, and doesn’t take into account the overall costs of sales.

Remember from the bake-sale, if you’ve got extra costs external to ad-spend (like $50 worth of primo brownie ingredients), your ROAS metric is never going to reflect true ROI – only somebody who is familiar with overall costs associated with a product or service can really interpret ROAS calculations for that product or service. ROAS is not the same as ROI.

As long as you keep it in perspective of larger ROI influences, ROAS as a guide to see how each variable affects profit margins can be quite interesting to look at. For those account managers that understand both their products and the metric well, there can be a benefit to being able to monitor ROAS at each level of a PPC campaign over time (and yes, ahem, tools are available). It’s often worth going the extra mile and doing some solid ROAS/ROI research before launching – it can help you get started off on the right foot with both the search engines, and your clients.

About Naoise Osborne

A former member of the Acquisio team, Naoise was the Search Engine Marketing Director at Catalytics, the services arm of Acquisio before that was sold off so that the company could focus more on developing the Acquisio platform. In his role at Catalytics, Naoise managed SEO strategy for agencies across North America. Today, Naoise is based in Montreal where he is a founding partner of AOD Media Group Inc., which specializes in SEO, analytics and Facebook marketing.

Comments

  1. Luke says:

    Nicce article, though I’d disagree and say there’s only one formula for return on ad spend: £ return / £ ad spend.

    The other formula you cited is a profitability indicator (neglecting cost of goods sold). In Google Analytics, the product of these equations are ROI and Margin respectively – different KPIs.

  2. Naoise Osborne says:

    I would tend to agree with you Luke, and in fact that’s how I set about writing this article. Then I went surfing around the PPC sphere and found that most sites (Bruce Clay for example and many others) seem to be calculating things with the profitability metric, so I thought I better write about both and help clear up the issue.

    Maybe I just made it muddier, who knows :)

  3. Ah wait, Luke you used the word ‘return’ without defining it… now I don’t know which formula you’re referring to. Ahh blessed confusion.

  4. Chouette! Comme il est intelligent, ce mec!

  5. Marc Poirier Marc Poirier says:

    David, how does Clix Marketing calculate ROAS? I think our readers would benefit from your input here.

  6. Gab Goldenberg says:

    Marc, a simplification occurred to me since I thought you were going there.

    In the revenue style metric, you have

    revenue / cost

    Where revenue = clicks x conversion rate x average order
    cost = clicks x cpc

    In other words, you have clicks ( (conversions x average order) /cpc )
    Or in English, clicks times all of [cpc divided by all of (conversions x average order)]

    Meaning that really, it’s just (conversions x average order)/cpc (since clicks/clicks = 1 and therefore doesn’t need to be expressed). Am I missing something?

    ROAS = Conversion rate x Average Order / CPC …

    ex.: 0.01 (a 1% conversion rate) x avg order $200 / $1 CPC = $2 ROAS – 2 bucks in revenue for each dollar in ads.

    Note: George Michie wrote a good post at RKG about how ROAS isn’t constant and that there’s a point where you have decreasing marginal ROAS so you eventually need to track the speed at which roas is changing. E.g. At the start you may get $2 ROAS, but then once you get a core part of the market, it takes more ad spend to gain marginal sales, so you need to figure out much more you can spend and still get a positive ROAS.

  7. Interesting article. I really like the example data that implies that a typial ROAS is in the 100% to 200% range. I think that if you also looked at including the Conversion Marketing Ad spend (personalized advertising to the shopper while in the website) then the metrics can be alot higher. We have seen consistent ROAS double for a site that combines an advertising focus of across getting traffic and then increasing the conversion rate with dynamic personalized offers. Just something to think about as more and more site also focus on the close after the browse.

  8. Maks says:

    Your second formula is for ROI, not for ROAS. it’s little different KPIs

Speak Your Mind

*