In the pre-BFCM blitz, questions about social ads are bound to come up.
Usually I urge caution when thinking about using organic social media content to drive brand awareness. Pumping out content ever faster is a quantity-over-quality race to the bottom. It’s the warped metric derived when someone over-correlates volume of inputs with value of outcomes.
For example, here’s an email I received, subject line boasting:
On the other hand,
PAID advertising on search and social platforms can be profitable.
Love ‘em or hate ‘em, SEM and social ads can drive positive returns for a campaign. At first, social ads seem pretty exciting: you put some money in, money comes back out. After while a question becomes obvious: how big is my profit on each campaign?
How to measure profitability of a campaign?
How exactly should return be measured for an ad campaign? instead of ‘ROI’ (which I don’t recommend anyway), we usually think in terms of ‘Return On Ad Spend’, or ROAS. It’s the relationship between ad spend for a campaign, and sales return attributable to the campaign.
ROAS = Revenue / Ad Spend
So..if we were to break even on a campaign, our Ad Spend would equal Revenue, right? Nope!
It is our Gross Profit, not revenue, that represents funds available for ad spend in our campaign. [And this is still oversimplified…we might want an allocation for other types of costs, but holding off on that for now.]
If you have only one product, life is easy. You can probably skip the rest of this article. If not, you will need a matrix of prices and volume for all goods sold. But you will also have variable cost of goods sold (COGS) for different products. COGS lets us calculate Gross Profit:
GP = Revenue - COGS
Using the ROAS metric above; the breakeven ROAS to use as a benchmark for a campaign is:
Breakeven ROAS = Revenue / GP
This is because Ad spend can be no larger than GP for the campaign to be profitable. If Ad Spend exceeds GP, you’re officially in the red. See the charts below.
Multiple, not margin.
Not only do we want to ad spend to be less than GP, but we want GP to be a large multiple of our ad spend. It is this multiple of GP / Ad Spend that we are hunting for; not a margin like Operating Income. The multiple is more useful for comparing different campaigns, which may have had different budgets.
Facebook Ads Manager Reporting is nice. The default ROAS calculation is misleading.
Facebook Ad Manager is very popular these days; we'll go with that for demonstration (we could apply this discussion to any number of ad platforms).
FB Manager allows us nice reporting views on campaign results. As long as we’re automatically feeding FB with sales results, we can review not only impressions, etc; but also financial performance of each campaign. Out of the box, FB Manager calculates ROAS. Great! This is what we want to use. However, FB’s calculation is a bit crude.
Naturally, Facebook Manager organizes results by campaign. Sales results are aggregated into a ROAS metric. Without any customization, FB will just report number of conversions on a campaign. In our formulas above, it only tells us be ’ROAS = Revenue / Ad Spend’. This is incomplete, as we need to know Breakeven ROAS.
We want FB to tell us what is the return incorporating Gross Profit on all sales, and ad spend. The real performance number we want to get at is: how many dollars of 'free cash flow' did each dollar of ad spend produce. Ideally, we could skip the benchmark comparison, and just say:
gpROAS = GP / Ad Spend
If you look at the charts below, you can easily judge performance of campaigns not to an internal benchmark, but can compare them to each other directly: is gpROAS better or worse than 1.0?
In this scenario:
- ROAS = 100 / 60 = 1.67
- Breakeven ROAS = 100 / 50 = 2
- 1.67 < 2 ☹︎
- gpROAS = 50 / 60 = 0.83 ☹︎
In this scenario:
- ROAS = 100 / 40 = 2.5
- Breakeven ROAS = 100 / 50 = 2
- 2.5 > 2 ☺︎
- gpROAS = 50 / 40 = 1.25 ☺︎
How to actually do this
To implement this approach, we need to incorporate COGS into the data feed to Facebook Manager for tracking on a per-order basis.
The path I would recommend is (may require a bit of programming in in your ecommerce platform, and FB):
- Incorporate conversion value into Facebook data; this will give the crude Revenue / Ad Spend metric (without COGS). You’ll want to add this functionality regardless.
- Add COGS tracking to your ecommerce site.
- Set up transfer of COGS data from your ecommerce platform into FB Ads Manager.
- Add a new custom property in Facebook to track COGS alongside other parameter object data tracked by Facebook’s standard events in a campaign. OR: If you’re ready to shift over to the gpROAS = GP /Ad Spend model, you can directly replace conversion value with GP in your data connection to Facebook. Then it will be easy to sort and compare campaigns directly in Facebook Manager, without any intermediary benchmarks.
This shift in reporting should drive toward more focus on to the most profitable products, market segments, and campaign strategies. You may still want to test some ideas outside of that as the only goal, which is fine. Let’s just not kid ourselves with the Revenue / Ad Spend metric as the final word on campaign profitability.