Jonathon Shaevitz is CEO of Maxifier. Here he talks about the Maxifier solution, managing yield on premium inventory and a real-life case study where the company's offering delivered for a European client.

Can you give an overview on the Maxifier tech offering in the European market?

Maxifier’s believes that all campaigns are performance campaigns. So while most of our customers are selling premium CPM campaigns, Maxifier improves the performance of all these campaigns, which in turn, increasing sell through rates and protects and grows margin. Maxifier’s initial offerings to Media companies comprises of two main areas; Campaign Performance Optimization and Advertising Intelligence Reporting. Together, they give Publishers, and particularly their Ad Ops teams, unprecedented inventory visibility and control through a single interface. Starting with pre-campaign simulations and continuing through the live campaign, our self-learning algorithms yield actionable recommendations that, in turn, eliminate costly development errors and maximize the performance of both the campaign and, ultimately, the entire site or network. We want publishers to know that “Every Campaign is a Performance Campaign” and, using our tools, to build “Campaign Performance Optimization” (CPO) metrics into everything they do.

How does adMAX optimize “premium” inventory? Can you give some insight into the process – and its uses by your publisher clients?

The Campaign Performance Optimization product within adMAX shows exactly what’s happening inside a campaign or set of campaigns the moment it is happening. Integrating data from any ad server, adMAX automatically displays everything from general campaign views to ultra-granular information in a single screen. It measures performance against defined business rules and gives the user the opportunity to take immediate action. Customers can quickly see where individual campaigns are over or under performing, use the information to make informed decisions, and shift ad placement to optimize performance based on a multitude of standard and user-defined metrics, such as revenue, CPM, CPC, CPA, CTR, and Brand Engagement.

You have positioned yourself as a premium product for optimizing premium inventory. Do you think there is too much focus on improving yield on unsold inventory – instead of focusing on the higher-margin ad impressions?

Absolutely! The 80/20 rule certainly applies when talking about general inventory vs. premium inventory. Not only are CPMs and the raw revenue far higher with respect to premium inventories, but premium inventories also have a significant impact on the value of publishers’ brands. When the ecosystem operates at less than full capacity, as is often the case, the results have been consistent: frustrated advertisers, publishers with unrecognized revenues, lost business. That’s why we have been laser-focused on optimizing premium inventories.

How would a publisher or ad network integrate adMAX? Is it platform agnostic?

Yes, adMax is designed from the ground up to be platform agnostic and will easily integrate with any publisher’s or ad network’s ad server. It normally takes only a couple of weeks to complete the integration process and start feeding data into the adMAX engine for processing and analysis. Furthermore, because adMAX is already in use with a variety of different publishers and ad networks using different types of servers, virtually all kinks in set-up process have already been worked out.

How does Maxifier work with ad nets and is it similar to the publisher solution? Also, how long does it take before adMAX collects statistics and starts generating suggestions?

There are some commonalities’ in Publisher and Network solutions. Clearly networks have different issues and are typically optimizing at the site level where direct publishers are optimizing at the page and placement level. For both, the key is to understand our customers’ business objectives and work flow, and in turn, improve their campaign performance and insight into their inventory.

In terms of timescale for data collection, with most mainstream ad servers, customers are up and running in a matter of days and experience immediate benefit. We then work with our customers over the next month or two to integrate other technologies and tune our algorithms to their unique requirements.

Can you give more detail on how Maxifier’s algorithms work in optimizing publisher inventory?

At our core, Maxifier is a mathematics company. We have dozens of different algorithms based upon different situations. adMAX is a self-learning system, which means that that our algorithms improve their performance over time. In all cases, Maxifier uses expert rules, mathematical statistics, heuristic learning algorithms and multi-agent systems to optimize publishers’ premium inventories. Our algorithms are incredibly accurate. Furthermore, the data we use depends on the type of analysis required. For short-term data analysis, the most recent two weeks are used. Long-term analysis, however, makes use of data from a longer time-period and incorporates specific historical, seasonal and other dynamic variables. In all cases, adMAX allows users full control over the systems recommendations. Users can have the system make automatic changes that clear a certain hurdle rate, or review each and every suggestion, its implications and payoff, before those recommendations are applied.

Maxifier delivers immediately actionable data to publishers on premium campaigns. How critical are performance analytics and inventory management in maximizing yield for publishers?

This has been the constant challenge facing the Publishers and Networks on and offline. The adage for media companies has always been: know your audience. Now they must add: optimize your performance. In order to thrive, our clients achieve optimal performance through a deep understanding of the interdependency of their inventory and how it can be managed for maximum ROI. With measurement and analytics tools provided by Maxifier, media companies get the opportunity to have a greater insight about the value of inventory and audience. It also provides the chance to demonstrate to agencies and clients alike the value of their environment and brand.

Can you set up the algorithm to act on this data without the need for manual optimization and a dedicated resource for trafficking – in the same way some DSP algorithms work for buyers?

While Maxifier can be set for automated actions and optimization, best practices advise in favour of human interaction and oversight because of the dynamic nature of sales, inventories and environmental vicissitudes. While it’s true that Maxifier forecasts are extremely accurate, depending on the availability of certain historical data patterns, it’s only through manual intervention and/or overrides that Maxifier would be aware of predictable spikes in impressions around a major forthcoming event (e.g The Olympics). With that being said, many clients choose to automate certain types of recommendations; for instance, if under delivery of a certain campaign is predicted and we can solve that problem without negatively impacting any other campaigns, that suggestion could be automatically applied.

If a publisher did decide to put premium inventory into a real-time exchange or SSP, could the Maxifier algorithm be deployed to deliver improved performance? If not, is this on the roadmap?

RTB is still very much in its infancy but growing steadily. With less than 40 billion impressions globally being transacted via RTB in this way, the Publishers need to ensure their core offering has higher sell through rates and that their teams get better at being able to demonstrate the merits of Campaign performance optimization as a first step!

Are there any case studies you can share about the success of the Maxifier offering in the European market?

The Guardian instigated the ground breaking use of Maxifier to optimise on engagement metrics. For Virgin Atlantic, this meant that guardian.co.uk outperformed the rest of their campaign for “Upper Class” by 64%.

The results:

All success metrics are evaluated from the initiation of the partnership in May 2010 to December 2010.

• CTR uplifts were monitored through both A/B (control vs optimised) testing and site-wide. A/B CTR uplifts averaged 33%. Site-wide CTR uplifts averaged 22%
• Global STR improved from 57% to 94%, an increase of 64%
• Direct Response revenues grew by 46% and DR eCPM by 35%
• Brand revenues also grew significantly, taking guardian.co.uk Y-O-Y revenue growth for the period to 64%
• Monthly average of optimisation changes 2016 across an average of 1200 campaigns
• Improvements in efficiency of inventory utilisation through reduction in under and over delivery created over £400,000 in value.