Key metrics to take into consideration when monetizing your website
When it comes to monetization of a web page I prefer to think of it as selling a user's attention. It’s a balancing act between offering the user the best possible product, while monetizing on the user’s intention. It is never sustainable to have one without the other for a business model which relies on advertising as its principal revenue stream. Without the product there would be no advertising, without the advertising there would be no product.
Going one step further, they also both compliment and negatively impact each other. The more interested the user is, the more likely they are to stay and carry on browsing. This means more pageviews, more engagement, and ultimately a higher chance of the user seeing and clicking on an ad.
On the side of monetisation the main focus should nearly always be Revenue per mille (RPM) and Value of the visit (VoV). This is fundamentally how much revenue you are generating for the traffic your page or site is receiving. RPM is typically used to calculate how much revenue you are generating from the pageviews on that specific page or pages, and is more focused on monetisation ability of the ad set up on those pages, while VoV goes another step further and takes into account the whole user’s session, and better analyses the flow of traffic.
RPM and VoV are the product of an eCPM and a fill rate, and pageviews per session. The eCPM is how much an advertiser is willing to spend for an impression while the fill rate is how many potential impressions you managed to sell. Again, these two metrics are directly related and it is impossible to touch one without it impacting the other, furthermore it's all too common of a practice for publishers to talk about these metrics in separate conversations. Knowing how far to push each one depends entirely on the demand, your back-fill options, and the price distribution of bids for that inventory. The pageviews per session are typically a result of the product being offered.
Whether it is advertising or any business, generally speaking any time a price is increased there will be a decrease in the quantity sold. On the other hand whenever a price is decreased there will generally be an increase in quantity sold. In advertising terms, this translated to eCPM and fill rates, and is the reason why we cannot focus on one without thinking about the other. It is natural to want to value our inventory at the highest possible level, but if we only sell 20% of our inventory with an eCPM 20 EUR, we are only making an RPM of 4 EUR. Equally, it is also normal to want to sell as much of our inventory as possible, but this will reduce competition for the impressions which in turn will reduce the value. Knowing just how far to push, and which value to have both of these metrics at to produce the highest RPM is the tricky part.
Fundamentally this is where the demand for the inventory comes into play, as well as your business’ strategy for pricing; would you rather favour maximising revenue or manageability of your inventory? If the answer is to maximise revenue then publishers should aim to have as granular a pricing structure as possible to try to stop advertisers from bid shading, and capture these often low by quantity, but very high yielding bids - in many cases the advantage to be had from pushing these high bids even higher is far more than that of the increased coverage.
There is a segment of inventory where 80% of the bids are in the 0.01 to 0.05 EUR eCPM price range bucket, 10% are paying 40 EUR eCPM, while the other 10% are somewhere in between. While in this example we will cut a large part of our inventory by setting a floor price of 40 EUR, the advantage to be had from pushing these advertisers from 40 EUR to 45 EUR for example is much more than if we were to win the other 90% of bids by reducing the price.
The downside of having this level of granularity however is that it is of course more time consuming as it can often be difficult to identify these peaks of demand, as well as the need to manage more rules due to higher level segmentation (e.g. geos, page type, position on page).
In contrast if manageability is the priority, while revenue will not be as maximised as it possibly could be, some of this will be compensated by the much increased coverage, as well as the additional free time you will have to work on other optimisations. In times of low demand however, expect advertisers’ buying algorithms to push down eCPMs to increase their ROI and try to exploit these broad targeted, and low eCPM floor prices.
On the other hand, partially thanks to the first-price auction model, in times of high demands it can sometimes be beneficial to have less restrictive pricing rules. Advertisers now have to pay the maximum bid, so whatever price they bid has to be beaten by another advertiser. This means that if there are a higher quantity of high-value bids, it becomes difficult for advertisers to bid shade as they will start to lose the impressions to others. As a result, we can enjoy low floor prices and an increased coverage, while not suffering on the side monetization, at least in the short term. A real risk that will present itself however is that over the long run, your sites’ inventory will likely be devalued by these bid-shading algorithms which can often be difficult to recover without hurting coverage
Example where low floor and broad targeting is favourable:
Whichever approach you choose as your pricing strategy, even more important is how to measure if something was a success or not and goes back to what I mentioned previously regarding RPM and VoV. Any change made should always relate back to these metrics as they directly tell you the revenue you are earning for the traffic you have. These metrics tell you the whole picture, and if you are confident in your pricing strategy and setup (e.g. not worrying about low fill rates or eCPMs just because they are low), then I'm confident you will yield more revenue from your site’s traffic.