You can't handle the truth...

The metrics that really matter in online gambling and the ones that really don't

Online gambling is a simple business, more often complicated by incompetence or technological overreach than underlying structural issues. You pay $X to acquire a player, they lose $Y to you and you just try and make sure your costs are lower than (Y - X). Find players, give them stuff to play, don’t do it all in on TV sponsorships and first class airfares. Easy game.

Yes it can get very, very, complex in the execution. Digital marketing alone can be a maze of internal and third-party tools, data points and projections with dashboards that look like something from 1980s science fiction guiding the steady hand of the fine people in marketing. But in some ways it’s as complicated as you want to make it.

I’ve known very successful mid-sized businesses doing CRM off a basic Excel spreadsheet with a small acquisition marketing team. Because ultimately it comes down to some very basic metrics and KPIs. If you’re on the right side of them you are going to make money, and if you are on the wrong side of them then the best office space in all of Malta is not going to help you.

What are you looking for?

But here’s the crazy thing: Most people outside of the core of the online gambling industry don’t really know what the important metrics are, or if they do they don’t know how to interpret them. We have an industry worth hundreds of billions where the majority of people investing in it, analysing it and even working in it don’t know how to measure success from it.

That’s not ideal is it?

So what should we be looking at and what do the operators consistently keep from view to stop us really knowing what is going on? Well let’s take a look at a few key terms and what they can and can’t tell us.

Handle/Turnover

Let’s get this one out of the way before we start. Turnover, the sum of every penny bet, is a lovely headline grabbing number but in terms of assessing the viability of a business it’s a chocolate teapot. Growth here can be a reasonable directional indicator, although this can be deceptive, but it only really begins to make sense when used alongside gross and net gaming revenue…

Hold/Gross Gaming Revenue (GGR)

This is Turnover minus Player Winnings. It’s all the money you get to keep once all the players have left the table, their bets have concluded or they have cashed out their winnings by whatever the end point is for that time period. A useful flash metric, but one that contains too much noise to be really useful.

Net Gaming Revenue (NGR)

NGR is GGR minus all bonusing and free bets. In other words all the imaginary money created by the operators and given out to the players that ends up being won back by the operators gets removed to give the true dollar amount won. So if you give a player a $50 bonus on a $100 deposit and he ends up losing all $150 back to you your GGR is $150 and your NGR is $100.

NGR is by far the best metric to judge performance. It can be combined with handle to give you a guide to win margin, which in turn can help you know if your performance is just due to positive or negative variance. If handle and GGR is vanity, NGR is sanity. And the basis for most of the other key metrics.

LTV

Lifetime value. How much a player will spend over the course of his customer lifetime, which is often defined as five years but can be shorter, longer or open-ended. The trouble with this is not all LTVs are created equally.

As one insider puts it..

I have never seen LTV calculated the same way twice. Including inside the same businesses between units, there can be huge variance

This is a key number that all operators try and estimate based on assumptions derived from historic data points from other customers as early as they can. If you know what the estimated LTV is you know how much you can happily spend on them in terms of bonusing, free bets and trips to the races.

It’s intimately linked to but not the same as…

ARPU

Average revenue per user. This is usually just a simple case of dividing NGR by the total active users for any specific time period. ARPU can be monthly , quarterly or annual but importantly quarterly won’t be 3x monthly and annual definitely won’t be 12x monthly because of a little thing called…

Churn

This is how many and how quickly your players say thanks but no thanks. This will generally be expressed in a few ways. First deposit churn, how many of your new acquisitions stick around to make another deposit, Month 1 churn, how many of your players last the first month, and then Month 3 and Month 12 churn. But you can slice it however you wish.

This is a really important metric and one that almost never gets disclosed. It shows how well the operator is managing to retain its expensively acquired player base, and this is vital because new players and old players are very very different.

FTDs

First time depositors. These are the fresh faces through the door. The excited new entrants to your online gambling world who want to see what you’ve got to offer and get some of that sweet, sweet bonus money. FTDs is a key metric in any business, as it speaks directly to revenue growth, but unless you manage to hold onto a good number of them beyond month 1 you will have problems.

Any FTD will likely be minimally profitable or often loss making, with the cost to acquire the player often much higher than the NGR prior to any second or third deposit. What we really want to see is….

Retained Players

Players who have been retained are those that stuck around to make a second deposit. It can also include players still working their way through a large first deposit, or winnings from early bets, but generally it means players who redeposited. Retained ARPU is normally » FTD ARPU and as a result this is a really key metric to know.

Some Nordic firms, notably LeoVegas, will disclose this but it’s pretty rare to see, more is the pity.

CPA

Cost Per Acquisition, or Cost Per Acquired Customer, is exactly what it says on the tin. How much money did you spend to get each customer through the door. But the thing is this can be a really crude metric or a really specific one and guess which one is more useful and which one is usually reported?

CPA that is just external marketing spend / number of customers acquired is helpful in giving a broad directional indication but what we really want to see is CPA (and in turn APRU) by channel. What we in fact normally see is entire marketing budget / total NGR for a marketing spend percentage.

This is helpful only as a like-for-like comparison with peer operators because there is so much hidden in marketing spend you struggle to get any clear picture of how that money is being spent. Marketing will naturally be a mix of acquisition and retention spend, and not all firms allocate spend to this line in the same way.

Actual CPA

CPA by channel (affiliate, PPC, digital display, broadcast, organic etc) is where you really start to get a handle on how the business is performing. Perhaps you are seeing fantastic low cost acquisition through a specific channel but significantly underperforming in others.

There is a lot to get into here and it is also interesting to see how businesses account for non trackable spend in CPA terms. This doesn’t just mean broadcast or outdoor advertising and sponsorships, but also headcount or how each business allocates promotions, bonusing, free bets and even affiliate spend within its marketing line.

In summary CPA and marketing spend numbers can hide a lot of secrets generally often confuse as much as they inform, so be very wary around them.

Cohorts

This is probably the most closely held secret of all. What are the cohort splits in the business. What percentage of revenue comes from the top 1% or spenders? From the top 5% of spenders? What percentage of players are depositing >$1000 a month, >$10,000 a month?

This may sound like it should be completely private information, but in a post regulation world it speaks to the sustainability of the business. A huge concentration of spend in the top 1% would not be a business I would rush to invest in based on the current environment.

And as a wiser head than me points out, it’s incredibly important for comparing businesses that may appear to be like-for-like based on the limited metrics we have. Averages can generate similar looking LTVs from a business with a stable, low-churn low-spend database and one where the business is 200,000 annual one deposit heroes and six blokes in Switzerland.

So what should we be looking for?

Well on a really basic level we need to know CPA, ARPU, NGR to have any idea of the operational efficiency and stability of a business. On a deeper level we need to know some kind of estimate of retention rates, some way of gauging player cohort concentration and a better oversight of marketing spend. But usually this is near impossible to get without an NDA.

Many of the largest gambling operators, particularly those focused on the US market, disclose very little of these data points and those they do disclose are often presented in a way that makes it hard to assess the underlying businesses. It’s not unknown to simply get given NGR, handle and an occasional glimpse at player numbers.

This may seem reasonable to you in terms of commercial needs, even though operators often share this type of data between themselves on an benchmarked basis via supplier reports. But without a deeper understanding of the metrics that really drive the business most analysts, investors and commentators are flying blind when it comes to judging online gambling performance.

And that, quite obviously, is a bit of a problem.