Metrics matter

The decline of the view-based-metric model has not been exaggerated. The quest to find an alternative is difficult. An overview.

The ad-budget crunch is felt across the media / publishing industry. While there are differences between markets, the overall trend is the same: publishers struggle generating enough revenue from the classic ad-based business models.

AdBlock and robots

One aspect of the overall situation is the growing number of AdBlock users. Now the company behind AdBlock Plus is even providing administrators with a solution to install their product across a whole network.

On the other side of the aisle, advertisers have to deal with the fact that a significant number of views on their display advertising is generated by robots and not potential customers. The Financial Times reported about a Mercedes-Benz campaign that had up to 57% of fraudulent impressions.

In a sample of 365,000 ad impressions brokered by Rocket Fuel over three weeks, Telemetry found that 57 per cent were “viewed” by automated computer programs rather than real people.

Current state of affairs

What is required is an industry wide acceptance and adoption of a new metric. Without a clear alternative in place, the view-based metric is creating the following problems:

  • For a few years already, there is a significant inflation in the value of display ads.
  • Ad space becoming cheaper forces publishers without alternative business models / revenue streams to accept ever worsening deals from advertisers and placing more ads on their digital properties.
  • This leads to even more devaluation of display ads.
  • On top of that, this trend aggravates the user experiences and frustrates the editorial side of the business. It destroys both the brand loyalty of the readers and the trustworthiness of the publication.
  • Furthermore, the waining cash reserves slash the chance of those businesses adopting new business models, because that would require substantial investments.

For some, native advertising seems like a natural way to alleviate those pains, but only a few publishers have adopted those into their core product yet. Compared with overall ad budgets, Native is nothing more than a niche in the business and doesn’t provide conclusive data and insights for the industry to adopt at scale. Additionally, most native-ad products are, as of now, far to expensive for most advertisers.

Engagement to the rescue

A new, industry-wide metric is required to stop the devaluation of advertising as a viable (main) business model.

Some publishers are lobbying for an engagement based metric model. Most prominently: the Financial Times, The Economist, Medium and Upworthy (which is radically transforming its business).

upworhty1-700x419 (Source: Upworthy)

Medium has been outspoken about their commitment to engagement-based models. Some authors are no longer being paid on the basis of how many times their articles have been seen, but how long people spent reading them.

For example, some people are now paid not by clicks, but by the total time spent reading in their collection—another experiment that could change as we learn what effects it has on the types of stories it helps produce, and how people find and read them. – Evan Hansen, Editorial Director at Medium

The engagement based metric would also reduces the strong dependency of most publishers on various social platforms and most of all on Facebook as sources for traffic. While receiving referrals will remain important, publishers will have the ability to optimize not only for click-bait, but for quality. Editorial teams will be able to focus again on the actual story.

If you let it, Facebook will take it all

As things are today, news publishing, especially in the US, is very much dependent on services controlled by technology companies. When Google closed down Google News in Spain, the traffic of the publishers who lobbied against the tech giant collapsed. This is a symbolic way of how the publishing industry is trying to cope with their current state. Instead of developing alternatives that would decrease their dependency, the strategic focus is far too often on blaming the tech industry.

Facebook has an unprecedented control over the digital distribution and spread of news, but the social network is far from satisfied being in control of how people discover news. In its everlasting quest to become the internet itself, Mark Zuckerberg’s company has started switching to a model that will reduce publishers to sheer content producers. This can be seen with their push into video. Videos hosted on Facebook perform far better than the ones who are only linked to / embedded.

“What the Shift to Video Means,” theoretically, is that much of the benefit publishers have derived from Facebook over the last three years, which required only occasional and modest adherence to Facebook’s explicit and implicit guidance, will disappear for organizations that are not interested in ceasing to be publishers to become “creators,” or in replicating their operations on another company’s platform just because it’s the momentarily dominant channel on hundreds of millions of new machines with poorly understood potential.

It makes a huge difference, especially on mobile, where Facebook dominance is now even stronger than on the web. On mobile videos hosted on Facebook will start playing (muted) automatically. What sounds like a small difference will create a distorted outcome for businesses that are operating with a view-based metric. Facebook is planing to adopt the same approach to non-video-based content as well. Publishers will be facing the decision of hosting their articles on Facebook itself, thus making it more likely for the content to be seen, but less likely that it will create any lasting value for themselves. A classic case of the middleman becoming rich and powerful enough to take over the whole stack.

Diversified business models are becoming non-optional for the publishing industry. There are alternatives to switching to different metrics or native advertising, but it is now clear that a continuation of the current state can only be considered wishful thinking.

In the end, it all comes down to Campbell’s law.

The more any quantitative social indicator (or even some qualitative indicator) is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.

Social Media Strategy: Analysis

How to measure performance and effect of a social media strategy to constantly improve it. This is part of our series about our social media strategy framework.

Measuring Success

You can only improve what you measure.

We like to work from data. To know if you’re successful, you should try to measure your activities against pre-defined metrics that match your goals. So let’s talk a bit about analytics & measurement.

Iterative Strategy Development

Digital communications are complex. So your strategies, actions, tools and campaigns should be permanently tested, analyzed and adapted.

The traditional model was to work from one big strategy through a test with a consumer focus group, and after some little adjustments to launch a big campaign. You’d measure the results and then start over. It’s a powerful process, but a slow, sluggish one. If it fails, it fails big. Chances are, it will.

Instead, we recommend working from more and smaller ideas. “Little Strategy”, David Armano calls it. You take this little strategy into an iterative cycle of planning, launch, constant measuring. The insights you generate along the way help you adjust while you keep going. It’s much more agile, in the sense that software developers use the word. You get to results faster, and if you fail it won’t hurt but rather strengthen you. At the core of this is constant analysis.

A scenario of future communications planning

We think that marketing managers will work much more like stock brokers in the future.

Instead of one big campaign that gets all the resources, they will have somewhere between 10 and 100 little initiatives going simultaneously, all with only a little budget behind them. Powerful analytics tools will help the manager to make real-time decisions about each initiative. Which one is performing well? Which one could do better with a little bit more energy behind it? Which one is clearly failing and can be canceled? Most of the budget will be in a big pool that the manager can draw from for individual initives that perform well. We think that this is most probably the only way to work the uber complex network of niche interests and nuanced target-groups. It’s almost impossible to predict the outcome of a big campaign so we have to change our approach to a thousand fires of small commmunications, investments and ideas.

Monitoring, Analytics, Reporting

When we say “analysis”, we use it as an umbrella term that spans monitoring, analytics and reporting.

Monitoring helps you listen outside your owned platforms.

Monitoring tools help you analyse the public conversations on Social Media platforms. Monitoring gives you a glimpse into your customers their sentiment towards your brand and products, their interests and needs. These tools are available from a wide variety of vendors, and you need to test which one fits your needs best.

Analytics help you listen on your owned platforms.

Analytics are strong in Social Media, and easily accessible. More than in most other other communication channels, engagement can be measured well here. Many platforms offer detailed measurements and analytics around users’ engagement with your content.

Reports help you make sense of what you’ve heard.

Reporting the results of Monitoring and Analytics is the basis on which to develop true insights, which will help you adapt and optimize your strategies.

Let’s expand on these a bit more.


Successful monitoring requires two things:

  • Clearly defined questions: To get to useful insights, you need to know what you want to know.
  • Establish processes: In order to make most out of the insights your monitoring generates, you need processes that make sure that these insights get to the right people and places, and can be implemented.

To give you a better feeling for monitoring, typical approaches may include:

  • Topic radar: A regular snapshot evaluates the online conversations around your key topics (retail, banking, design…).
  • Opportunities for engagement: Constant monitoring identifies opportunities to engage with users around the topics you’re an established expert in. This could mean answering questions about these topics in general, or about your brand in particular.
  • Feedback: Expand on the regular snapshots to include opinions about your brand or products, and based on this develop concrete recommendations for further action.

About the monitoring industry

We’re still in the infancy of the vendor industry around social media monitoring. Most monitoring tools available out there are still mostly a display of their technical capabilities but fail to act as a true source of insights. This has lead to most monitoring strategies being driven by whatever the chosen vendor can offer, instead of developing a clear strategy of what insights a company wants to see before choosing the vendor.


There is no single ROI for social media.

–Richard Binhammer, Dell Inc.

How to measure Social Media? The discussion around the best Key Performance Indicators (KPIs) has been going strong for quite a while now. While no universally accepted KPIs have emerged as of yet, there is a growing consensus on the basic approach:

  1. Define clear goals that match your business processes and metrics.
  2. Regularly adapt your KPIs and metrics according to new insights and industry consensus.

Altimeter Group developed a measurement framework that helps define the metrics you need. Imagine a pyramid that is made of three layers. Top to bottom: Business Objective, Business Metric, Social Media Metric.

Business Objectives are the clearly defined objectives that a company or department strives for. Typical business objectives include increasing sales, exploring a new target audience or increasing brand awareness.

Business Metric: To make the business objectives measurable, metrics are defined for every one of these objectives. The metrics could come in different flavors ranging from sentiment analysis to lead generation or cost reduction. Initially, each metric is watched closely. Based on the insights, objectives are defined around these metrics over a certain period of time.

Social Media Metric: At the lowest level, social media metrics are measured. They can show quickly which activities lead to good results. Typical social media metrics include Likes, fans, followers, views etc.

This brings us back to the first part of the social media strategy framework: the definition of business objectives. The set of metrics chosen should be as close to the business metrics as possible to make sure that the social media engagement keeps on track to achieve those goals.


Just having data won’t help much to improve a strategy and engagement. We have seen companies drowning in data collected from the social web who, having no idea what to do with all the information, just stopped collecting.

Regular reports can be a good way to take all the data collected from monitoring and analytics and analyze them for learnings and insights, which then can be applied to the company and the strategy and turned into actionable recommendations. This a part of the strategy that profits greatly from pure experience. Pattern recognition capabilities and a vast pool of knowledge are needed to look at data and draw the right conclusions from them.

To give you an example, reports could be structured like this:

  • Monthly: Social Media Metrics Report
    An analysis of last month’s social media metrics and overview of the monitoring results. Recommendations for concrete activities in the daily social media engagements.
  • Every three months: Business Metrics Report
    Analysis of business metrics and recommendations for optimizing the social media strategy for short-term and mid-term objectives.
  • Every six months: Business Objectives Report
    Analysis of the progress towards business objectives based on the ongoing reports. An executive summary of key findings and recommendations for the top level management. If applicable, a revision of the mid-term and long-term objectives.

No report will change anything if it’s only glanced at and thrown into a drawer. Reports are there to improve the strategy and that’s how they should be built: with a complete focus on the meaning of results and clear actionable next steps.

A strategy that is iterated constantly stays current and flexible. It adapts to new objectives, target-groups and platforms on-the-fly. And it will never be outdated. But in the end, its success comes down to a different aspect: how it’s implemented in the company. That’s what we will look at for our final installment about our social media strategy framework.

The next article in this series is: Social Media Strategy: Internal Organization.

What we read this week (23 Sep)

It was a week of light reading for us: Robot clocks, meaningful metrics and leadership by serendipity. You know, just the usual. Enjoy your reading.

It was a week of light reading for us. Here’s what we got for you.

  • BERG London: Product sketch: Clocks for Robots
    BERG in London has been thinking quite a bit about how to make the world robot-readable. Here are some thoughts (and sketches) on how time & place change, and can be made more easy to navigate, for robots. Meet the robot clock.
  • Love Metrics – Only Dead Fish
    Neil Perkin shares his thoughts on meaningful (as opposed to vanity) metrics.
  • Joi Ito: Thoughts on leadership
    “The cost of planning, predicting and managing rapidly changing, complex systems often exceeds the cost of actually doing whatever is being planned and managed.” Joi Ito on leadership in an age of decentralization, disruption and networked innovation. Embrace serendipity!

You can see all our reading recommendations (including the archives) at “What we read this week“.