Marketing ROI Calculations (Patreon)

Alright, let's talk about Return on Investment for promotional spending. I'm going to look specifically at advertising / promotional cost, and discuss why it's difficult to know what is working and not, and methods to get around that.

Defining Return on Investment

Firstly, let's define ROI. It's pretty basic in terms of the equation which is =  return (i.e. total income generated from the advertising) / investment (i.e. total advertising spend).

Let's give an example using Amazon Marketing Services (AMS) ads.

Total return = $100 (i.e. new income generated)

Total Investment  = $10 (i.e. spent on advertising keywords).

ROI = 100/10 = 10.

In AMS, this is the number you'd see as ACOS (sort of. ACOS in Amazon takes total revenue Amazon earns, not what you earn. So you have to calculate YOUR income which is often 70% of that amount - or less - and also add KENP numbers back in for your actual return).

Simple(ish), no?

No? It gets worst. Hold on to your chairs.

Tracking online - the basics

Now that we've discussed how to calculate ROI, let's talk about how we track sales online. For most software, we track direct sales via what are known as cookies - small bits of code that live on a browser and follow you around from browser to browser, reading what you do.

These cookies happen or register an incident when you take specific actions.

For example, with an AMS ad, it registers you clicking through on that advertisement at that moment. It then keeps track of what you do for the next 14 days (AMS's preset limit) and if you do buy the book, it finally registers the sale.

Yay!

So, with AMS, you have up to 14 days to register a sale.

With Amazon's Affiliate marketing, you have 1 day to register a sale and get the money from it.

With Facebook pixels and advertising, you can NEVER register a sale (because Amazon doesn't let you into their checkout so you can't tell if someone completes a purchase).

If you had your own ecommerce website, you could potentially track a sale from Facebook all the way to a final sale. Or any other 'event' (registering for a newsletter, etc.).

There are additional complications.

- some browsers reject cookies automatically

- some individuals 'clean' their browsers regularly, so that your cookies never register sales

- often, these cookies only track 'last click', so if you had clicked on the ad 3 previous times, it only registers the fact that you saw and used 1 ad. Even though it was because those 2 previous times helped convince you to buy it.

- ads that generate awareness (i.e. are present for people to see) but are not clicked on can never be included in the above sales calculations.

And digital marketing is THE best way to track sales. There are other methods to track digital sales (using unique telephone numbers, unique sales prices or landing pages, tracking from server logs, etc); but for our case, cookies are mostly what well worry about.

If you do brand marketing at all (non-digital marketing, display ads, etc.); calculating direct ROI becomes even more difficult and obscure (and somewhat pointless since the focus is on brand building / awareness which requires market research surveys rather than actual sales estimations).

Anyway, you see the issue with ROI tracking for your advertising spend.

Baseline and Estimates.

So what do you? You create baselines graphs of your sales and from there, you can calculate what the effect of each advertising campaign has on your baseline sales.

For example:

You sell $1000 worth of books a month

You then spend $200 in Facebook advertising that month

Your new revenue is $1400 for that month where you're advertising.

As such, your marketing ROI is $1400-1000/200 = 2x

That's it.

Simple, yes?

The Complications

Here's where things get complicated. Firstly, it's getting what you could consider a 'regular' baseline. Sales for books drop after month 1 & 2, often on an almost consistent basis. So, while you might sell say $100 in month 3, in month 4 (without any marketing), your sales would be $90. And month 5 might be at $75...

To calculate a proper baseline, you'd need to take the drop in these last 3 months and draw a trendline for month 6 (where you start advertising). Which means you have to be willing to sit back and watch your book sales (or heck, your entire series sales) drop for 3 months before you do anything.

If you release fast, then you might never have 6 months of uninterrupted data since a release in month 5 of book 2 would push up sales for book 1 again, meaning that you'd have to reset.

So you have to use an estimate of what book sales would be, often with only a few months worth of information. 

Secondly, you have to calculate everything based off individual series for income and sales, but still be cognizant of sales rollover from other series. What I mean by that is, if you release a book in another series in month 3, you might still see some sales increase for your first (unreleased) series because some fans might move over to check out your other work. So that will shift your 'baseline' numbers.

Lastly, it's worth understanding that some forms of understanding (arguably all) takes months to build up. That old saw of people needing to see your advertisement 7/11/13 times to make a decision holds true.

And in fact, there's sometimes a logramithic effect to repeated viewings, where an advertisement that started in month 1 that does little begins to have an oversized effect by month 3.

That, of course, depends on the kind of advertisement and when it applies to your readers/customers in their sales funnel.

Synergistic Effects

One last thing, talking about sales funnels. There's often a synergistic effect to marketing types. Brand building by itself won't affect your sales hugely (public relations, blog posts, Facebook takeovers, social media discussions, etc.). 

Brand building with AMS might see a significant increase, since you are now touching them at the top and bottom of the chain. If you find a way to reach them in the middle (Facebook maybe?); you'll see much more effect then.

Of course, you have to make sure you're 'touching' the same people each time, rather than doing PR to one group and AMS ads to another, which requires you to know your target market well.

Promotional Calendars.

So, with all those complications, how do you track baselines and do this right? You create a promotional calendar.

You figure out exactly what your current baseline is (based off whatever data you have, whether it's 3 months or 6 months). Preferably on a series that you aren't pushing too much. 

Then, from that baseline, you build your promotional calendar. You start by adding each type of advertising you intend to do. You give each of those a budget and a certain time period, before you add the next promotional campaign (so that you have a new baseline). 

For example:

- month 1 -3 - baseline stats created and estimated for month 4 & 5

- month 4 & 5 - you start advertising via AMS. You track revenue increase / change between months 4 & 5 to your baseline. 

You also create a new baseline with AMS ads using month 4 & 5 sales data for months 6 & 7.

- month 6 & 7 - you now stop touching and adjusting AMS advertisements. You leave it as it stands and add a new advertising method. Let's call it Facebook advertising.

You now track your new revenue numbers for months 6 & 7 and compare to the baseline statistics you created above to see if there's a resulting increase.

And so on, so forth. Personally, I might give it 3 months rather than 2, both to get a better baseline and also to decide what works best and allow me to test.

I might also just test 1 format (facebook ads) over a series of different audiences and format for months on end before I decide to move on and add something else like AMS. Or vice versa.

All this is a lot easier to learn best practices and what works when you have multiple series.

Hopefully this clarifies marketing ROI tracking and how best to get the most from your advertising dollars.