Why marketers need to park the CPA model and work with dCPM on programmatic


The modern marketer’s predicament

Marketers no longer carve out budgets from one very big number – those days are now long gone. Today’s marketers work with targets adjusted by channel that may change in real-time. Measurability has turned Digital marketing into a sales function: marketers are accountable and are under pressure to deliver revenue and profitability. But problems arise when the marketing models are over-simplified and don’t serve the purpose that they were designed for. For example, the user acquisition goal should be to acquire high-quality new customers at a cost lower than their Customer Lifetime Value (CLV) . Both the CLV and the acquisition cost are variables. However, many advertisers are mistakenly translating this objective into a strategy to acquire all new customers at a fixed CPA (Cost Per Acquisition).

CPA is seductive

Many marketers buy at a fixed CPA rate. This enables them to control advertising spend and new customer targets. But more importantly CPA is seen (erroneously, as we will see) as risk free because the advertiser only pays for newly acquired customers. In other words, they view paying CPA as a win-win scenario because they are precisely managing their spend and delivering results.

All is not as it seems

The underlying assumption of CPA is that all users are of equal value. This is evidently flawed. This way of thinking does not align well with programmatic and focuses on the publisher and not on audiences. In addition, as we will show below, there are other problems too.

The challenges of scale and quality

The CPA model works by paying a flat fee on acquisition, no matter which customer was acquired. For publishers running CPA campaigns that means that most ads displayed will have zero value to them. Furthermore, for publishers with low click-to-conversion ratios, there’s a higher risk of generating no revenue at all and that translates into zero incentive. Therefore, the CPA model places 1) a limitation on the number of publishers willing to accept your campaign, 2) a limitation on the scale and reach that your campaign can achieve and 3) a limitation on the quality of publishers willing to accept your campaign – the majority of premium publishers which generally speaking attract high-quality users may have a high media cost but will deliver customers that have a higher customer lifetime value.

All consumers are not equal in value

As we’ve already discussed, the CPA model assumes all users are of equal value and conflicts and places a constraint on the acquisition goal of finding high-quality new customers. When a high-quality user (predicted by past behaviour and interests) visits an app, the advertiser should be bidding higher to win that ad space and place an ad in front of that user.

A race to the bottom

The paradox with working with fixed CPA models is that it tends to acquire lower and lower quality users and make acquisition strategies fail. As a marketing manager calculates a theoretical Customer Lifetime Value (CLV) and starts a campaign at a fixed CPA, the campaign will provide the campaign will provide not only good, but also bad quality customers. As a result, the CLV may be lower than initially hoped for due to the mix of good and bad customers, resulting in the manager reducing the CPA target. In return, this causes the blend of newly acquired customers to worsen, which results in an even lower CLV and a poor acquisition strategy.

By contrast, good strategies focus on solid and trusted partnerships between advertisers and programmatic buying platforms. Trust which comes in the form of transparency from the platform in return for access to valuable data from the advertiser. As a result, both parties can cooperate in a buying strategy that goes beyond achieving a CPA, but which focuses on acquiring customers with a Customer Lifetime Value higher than their acquisition cost.

This not only results in a successful acquisition strategy, but also enables the programmatic platform to achieve the reach that the advertiser desires, without eliminating quality publishers just for their media cost.

dCPM at the heart of good strategy

Enter dCPM (Dynamic CPM), an advertising model that determines in real-time the value for each individual impression based on many factors and takes into account the performance goal for the campaign (and not just the cost of the acquisition). dCPM addresses the problems with CPA and CPM pricing when working with performance goals. At end of the day, an advertiser should achieve the campaign objective of acquiring customers at a lower cost than their customer lifetime value, and they should achieve this by paying a fair price for the media based on many parameters that have come into play to determine the value for a specific ad placement


In order to leverage the rich data set offered by programmatic and focus on high-quality new users, programmatic offers a good solution with Dynamic CPM that gives marketers transparency, insights and optimization options for improving campaign performance.


Adding more VAST, MRAID, NATIVE and PRE-ROLL mobile video inventory

mobile video

There’s increasing demand for mobile video ad inventory from both performance and brand advertisers. That’s good news for everyone, because more engaging ad formats generate greater benefits for all parties in the value chain: advertisers get more brand engagement from consumers and publishers get higher CPMs from video.

We are expanding our mobile video enabled ad inventory across ad exchanges for VAST, MRAID, Native and PRE-ROLL. Being able to support multiple video ad formats gives advertisers scale across publishers and exchanges and options for optimization. Across all ad exchanges we can now access over 10 billion video enable bids per month.

Supply is becoming a problem as pointed by the CEO of AppNexus in a recent article: the challenge and critical success factors for demand players will be getting access to global video supply  to support both performance and branding campaigns.

VAST mobile video

First and foremost, VAST is only available at scale on a number of ad exchanges. Secondly, VAST has not been as widely adopted as MRAID by publishers. For that reason, there’s limited non-skippable VAST inventory out there – therefore it follows that you can expect to pay a higher price. VAST skippable inventory has been more popular with publishers as they balance the trade-off between user experience (forcing users to watch an entire video) and monetization (higher CPMs).

The duration options of a VAST video will depend on whether it’s skippable or non-skippable. One of the drawbacks of VAST it that it only supports basic end cards – no carousels.

On the flip side, VAST is the established standard for video and integration between publisher, exchanges and demand players is relative straightforward. For ad ops teams, VAST tags hold all the attributes about a video campaign and that means detailed reporting and third party ad tag support. VAST works across mobile sites and applications.

MRAID mobile video

MRAID is the only viable option today to achieve massive scale across ad exchanges. MRAID has already been established as the standard for rich media since version 1 was released in 2011 and the updated version 2 in 2012. Videos can run in any MRAID compliant container that supports HTML5. For brands this particularly exciting because it gives them a white canvas to mix rich media interactive campaigns with videos in any combination or style. MRAID works across apps only.

Native mobile video

Facebook, Snapchat, Instagram and many more have already proven that native mobile video ads outperform other mobile video ad formats and brands are allocating advertising dollars to advertise on these apps. Smadex will be integrating with ad exchanges and direct publishers that support native video over the coming months.

To find out more about running mobile video campaigns with Smadex contact sales[at]smadex[dot]com

Smadex hires ex-Adfonic co-founder Paul Childs as Chief Revenue Officer

Paul Childs

We’re excited to announce that mobile advertising veteran Paul Childs has joined us as Chief Revenue Officer, where he will be responsible for marketing, business development, revenue and repositioning the business to accelerate global growth.

Paul was co-founder and COO at Adfonic (now Byyd) where he was instrumental in building the most successful mobile ad tech firm in EMEA. After Adfonic, Paul was VP GM EMEA at MoPub where he established business operations in the region and more recently CMO at LoopMe, where he helped them establish a leading global position as a mobile video DSP. Paul has also held a number of advisory board roles in the mobile ad tech space.

Commenting on his new role, Paul said: “The key success in ad tech is threefold – focus, hire super smart people and build awesome technology; because ad tech is becoming extremely complex and you need people with engineering or data science backgrounds to succeed – that’s why every C Level person at Smadex has an engineering background. There are not many businesses out there that can process hundreds of thousands of global transactions per second. Smadex aims to scale worldwide and support thousands of customers with their mobile marketing objectives.”