Reacher’s unique pricing logic — how it gives Mittmedia higher prices for ads AND advertising customers more value for money
During recent years, MittMedia has brought innovation and success to its digital advertising business. Our journey towards innovation and success has consisted of many steps along the way, and is the product of many people’s hard work. In a recent blog post my colleagues, Håkan Hamrin and Henrik Mazzanti, explain some of the work involved in purposefully increasing our digital display ad revenues, in spite of a reduced advertising inventory. In this post, I will delve deeper into one of the central success factors. I’m referring to the unique pricing logic that we have built into MittMedia’s sales platform Reacher. It is a logic that simultaneously provides MittMedia with higher ad prices per viewing sold AND gives our advertising customers more value for money from the same number of ad servings and campaigns bought.
In their post, Hamrin and Mazzanti describe how MittMedia has been extremely successful in increasing revenues on manually sold display advertising through the marketing tool Reacher Marketing, MittMedia’s sales tool for effective marketing.
Reacher’s in-built and completely unique pricing logic gives BOTH the advertising customer more impact from each display campaign AND a doubling in pricing for MittMedia from the same campaign.
This may perhaps sound like nonsense.
An impossible bit of magic.
In fact, it is completely and utterly true.
In this blog post, I describe how the logic behind our success works, and how we arrived at it, step by step. Five years of solid work have laid the foundation for the results we are now enjoying, and I will guide you through the insights, mistakes and successes we have experienced along the way.
The starting point in 2014 — comprehensive work on our strategy
In 2014, MittMedia was in the midst of shaping our future digital strategy, which also applied to the advertising business.
The fundamental questions we sought the answers to are very common in standard strategy work, but nevertheless central:
- Where are we today?
- Where do we need to be in the future and when do we need to be there?
- How do we get where we want to be, and in time?
Although the strategy we needed to shape applied solely to the future DIGITAL advertising business, the time frame and speed of development depended on the status and development of our print advertising business.
In 2019, the print business is still a large part of the MittMedia business.
In 2014, it was gigantic compared to the digital business.
This hardly gave us much confidence though, as the print part of the business was, and had been for many years, on the way down.
Expected decline in print — our first premise
When we sat down to work through our strategy, the forecast decline in print was also a driving premise.
As a baseline, we chose to use a pretty tough as-is scenario until 2020.
Based on the negative trend curve and projected decline in print, we were able to take the next step and draw the curve of the growth we needed to generate in the digital business to offset the loss in print. We chose to sketch out that need entirely based on the display business, as this was the only part of the digital business that existed in 2014.
Revenues needed to offset print by 2020
By drawing the curve we were able to visualise the business impact needed from the display business by 2020 to offset the loss in print.
Note: None of the graphs in this post are authentic with exact values, they are solely intended to illustrate principles and approaches.
Once we had drawn the two curves we had our answers to the first two questions in the shaping of the strategy: “Where are we today?” and “Where do we need to be in the future and when do we need to be there?”
Could the existing reach model get us to where we needed to go?
It was time to dig deeper into question number three: “How do we get where we want to be?” First, we looked for the answer in the existing business model for display, i.e. the traditional reach model we worked with in 2014. Would we be able to achieve the necessary future digital business need using the existing pricing and business model in the display business i.e. the reach model?
Was there any chance of us achieving our business goals in the display business by 2020 if we continued to use the same reach model?
We re-calculated the curve above in terms of the number of locally sold ad exposures needed per week by 2020. Then we looked at:
1. Number of ad exposures per page view.
2. Best practice in the industry for page views per visit.
3. Converted this into the number of weekly visits needed.
4. Best practice in the industry for the number of unique visits per week.
5. Converted this into the number of unique visits needed per week.
The result we landed at indicated that we needed 130% of the local digital reach per week.
To manage this we would have needed a large immigration of individuals into MittMedia regions (a 30% population increase) and that all of these new arrivals would choose to be frequent MittMedia readers, every week.
In theory a desirable scenario, but completely unrealistic.
This meant that we had to do more. For example, we could:
- Increase the number of ad exposures per page view.
2. Remove the lazy load for ad servings that we had just developed.
These two measures would in turn make both the reader and user experience practically intolerable because of all the ad servings, and also we would have to charge for ads that were not visible. It was the complete opposite to the focus on ad quality we had spent the last few years developing.
Can we get to where we want to go with higher prices within the existing model?
The above option, trying to maximise the existing business model through drastically increased reach or increased ad volumes was not relevant. Especially not in the local markets we operate in, and given the credibility in news reporting we have to deliver.
Consequently, the next issue was to look at how we could increase the value/pricing from the reach and advertising inventory we actually had.
We asked ourselves what increase in value from each individual ad exposure would be necessary in order to reach the impact and revenue needed from the display business in 2020, to compensate for the loss in print.
It turned out that in 2020 we would need an average of five times more in value than in 2014.
This in turn meant that if we retained the same advertising model, we would have to dramatically increase pricing in the CPM levels and present price lists to our customers that would make it impossible to do business. We had to do something other than simply increase CPM levels and retain the same business model.
Need for dramatically increased advertising prices towards 2020
Conclusion: The reach business had to be replaced by a target group model in which ad servings were enriched with data.
We needed to find a way to add much more value to the ad exposures so that we could deliver higher value/impact to the customer and thereby justify a higher price.
Our theory was that we would be able to achieve this if we added data to our ad servings and offered customers more effective target group advertising
Our first step was to build a pyramid containing the types of ads known at the time, with the lowest value at the bottom and so on upwards. We simultaneously looked at how much inventory needed to be used for each campaign. We overlaid this pyramid with the revenue need curve, now converted into a CPM value.
Price increases by switching to a target group model with data-enriched ad servings
The pyramid and the curve combined made us realise that this was, in theory, doable. Very challenging and difficult, but doable.
The question was HOW we would simultaneously manage to:
- Get a higher value on each individual ad serving by enriching it with data.
- Make the customer themselves want to pay more.
- Steer our business away from a deeply rooted discount culture.
- Optimise our inventory.
A hefty challenge, and to make all parts of it work we had to do a range of different things:
The first step: Collect data and enrich our advertising inventory with the data
Most of the work was obviously in creating a digital eco-system allowing us to collect, process and add user and customer data to our ad inventory. We also had to create and deliver a new advertising model built on target group advertising instead of a traditional reach model, through a simple and creative tool (Reacher).
It was important, though, that when we charged for the above inventory in our tool according to the new model, we created a coherent logic in our pricing.
Next step: Coherent, logical pricing
This is where we landed in the Reacher pricing logic, that I will now try to explain through some examples:
In the examples below we have made some assumptions:
- We have set a base value CPM of 100 in our site network.
- We have a total inventory of 10 visitors.
- We have a frequency of 1 per visitor.
The next step is to see how we could build a logic that gives higher prices AND greater customer benefit if we shape three different example campaigns, in which we gradually enrich ad exposures with data.
- ROS campaign.
- Campaign steered by enriched geographical data.
- Campaign steered by enriched geographical data and target group data.
Reacher’s pricing logic — how it works
The pricing logic works in such a way that it starts from the entire inventory.
After the advertiser’s choice of geography and target group, the pricing logic does not use the inventory actually chosen, but the inventory that has been deselected, that is, those visitors that the advertiser does not want to reach with the campaign.
From a technical perspective the pricing logic is built on a unique algorithm in the Reacher platform. The algorithm is obviously more complicated than in the example, but the fundamental principle is the same. This is how it works:
Campaign 1 — ROS
In this campaign we use 100% of the total inventory. The CPM factor is thus equal to 1. This means:
- The CPM value is 100 (base) multiplied by 1 (factor) = 100
- The cost of the campaign for the customer therefore becomes SEK 0.10 (CPM value) multiplied by 10 (visitors) = SEK 1.
Campaign 2 — Steered by enriched geographical data
In this campaign 40% of the total inventory is NOT used, making the CPM factor = 1.4. This means:
- The CPM value is 100 (base) multiplied by 1.4 (factor) = 140
- The cost of the campaign for the customer therefore becomes SEK 0.14 (CPM value) multiplied by 6 (visitors) = SEK 0.84.
Campaign 3 — Steered by enriched geographical data and target group data.
In this campaign 80% of the total inventory is NOT used, making the CPM factor = 1.8. This means:
- The CPM value is 100 (base) multiplied by 1.8 (factor) = 180
- The cost of the campaign for the customer therefore becomes SEK 0.18 (CPM value) multiplied by 2 (visitors) = SEK 0.36.
The result: More value for BOTH advertiser AND MittMedia
What is the difference for the advertising customer and MittMedia using Reacher if we compare campaign 1 (no data or target group) and campaign 3 (with data and dual target groups)? Do we obtain the price increase potential that we wanted to push in the pyramid shown earlier? Yes actually it looks very good indeed.
For the advertiser:
- There is a clear financial incentive to create campaigns directed to smaller target groups.
- It allows you to create several unique campaigns for the same budget. In this example the advertiser can run three different campaigns tailored to three distinct target groups, for the same campaign budget as for 1 ROS.
For MittMedia or other media groups using Reacher:
- The model automatically drives an increased CPM value per ad serving.
- MittMeda gets a less-used inventory for the same revenue per customer and can sell the remaining inventory to other customers. If the customer chooses to use their entire budget and run three campaigns according to campaign 3 above, the media group retains 40% of the ad inventory to sell to other customers. (Two visitors per campaign multiplied by three campaigns = 6).
The conclusion is that, as unlikely as it may seem, we can actually achieve our goals:
By enriching our ad inventory with data and also constructing this pricing logic, we can obtain the pricing needed to reach the strategic goal related to the impact of the display business revenues offsetting the loss in print by 2020.
At the same time as we increase pricing, we are also able to give advertisers more bang for the buck, even in the higher price brackets.
The pricing logic in combination with growing campaign volumes is fuelling success in terms of total revenues
As we have built this unique pricing logic in our sales tool Reacher, we will have the ability to deliver the effects of the logic and the data enriched advertising inventory into each and every campaign that our sales people sell to their customers, every day.
Higher pricing and the growing campaign volumes will deliver the revenue increases that we need to take us towards the strategic goal we discussed at the beginning of this post.
Straightforward, right?
Yes, absolutely, but it has not been quite that simple.
Sure, MittMedia has achieved record-breaking price and revenue increases in the display business, both from a Swedish and international point of view. But we are not yet entirely where we needed to be ahead of 2020 if we look back at the “offset curve” mentioned earlier in the post. One of the reasons for this is that we have not yet completely opened the floodgates for demographic data (the top of the pyramid above), for different reasons, but we should probably have made more progress by now anyway.
The question is why.
New business models demand new governance models
Looking closer at the numbers, we realise that just after launch we saw an effect that we had not counted on, illustrated by the graph below.
- The CPM value and revenues per campaign increased dramatically, as an effect of the strategy and the logic above.
- However, total revenues did not increase, instead they remained at the former level.
- The number of orders in Reacher using the new logic fell in line with the increases from the CPM/campaign value effect.
And this was not the intention of course. Wasn’t the total revenue supposed to follow the curve of the CPM value and revenue per campaign?
What happened?
The reason is actually very logical and something that we should have foreseen. The culprit was business governance: With all due respect to budgets, this time they caused problems.
Our sales people continued to focus on reaching their budget, and not on steadily increasing the number of campaigns with higher pricing in Reacher. This meant that the number of campaigns sold in Reacher and with Reacher’s pricing logic actually fell.
Or, if I relate it to every day work:
The higher pricing in Reacher helps sales people achieve their budgets by selling fewer display campaigns. Once sales people reach their budget, they turn their focus to other parts of the product portfolio.
The number of display campaigns sold in Reacher with the improved pricing logic thereby falls, and the total revenue curve deviates from the curve for CPM value and revenue per campaign.
Our conclusion was that the fault was not with the logic or technology, but with the fact that new models demand new governance models and working practices for the people that conduct everyday business.
Our advantage: We own the problem ourselves and can work on solving it
The good thing about the challenges identified in business governance was that we owned the problem ourselves. For MittMedia this meant re-working the KPI’s to put a clear focus on activities that drive progress in the right direction.
In this case, it largely involved changing the culture, and KPI’s such as “Active sales people in Reacher” and “Number of purchasing digital customers per week” helped us to change how things developed.
The impact: Reacher, together with changed business governance, led to the local display business growing by as much as 28 percent in 2018.
Last but not least, I am proud to tell you that of MittMedia’s five (!) nominations for the INMA Global Media Award 2019, which takes place in New York in May, Reacher has been nominated in two categories!
Joaquim Linder, Product Owner Reacher MittMedia
You can reach him at: joaquim.linder@mittmedia.se