Previously, we explained how to plan for successful audience activation before a content premiere. However, while planning is essential to success, it is only the start. Campaign tracking and post campaign attribution are required ingredients for any marketing campaign.

In this article, we will take you through how demand data offers a powerful way to perform both campaign tracking and post-campaign analysis, tailored to the needs of marketing in the attention economy.

Demand offers real-time Campaign Tracking

As any marketing professional knows, unexpected events can impact any campaign. Even if you have a great target profile, know which channels to use and when to deploy resources/assets, it is essential to have access to real-time feedback during the campaign.

Demand is an ideal metric for campaign tracking. Unlike older tracking technologies, demand inherently captures multiple consumer touchpoints like social, research, piracy and social video. These platforms are where modern audiences signal their desire to consume content. More than that, demand combines these signals weighted by their importance as a signal. Events like targeted video ads, a hashtag that goes viral or an ongoing TV advertisement campaign all have different impacts on demand that indicate the consumer journey. This means that once disparate and siloed signals are combined into a powerful measure that can be used to identify which channels influenced the consumer to watch this particular show and attribute accordingly.

As the demand metric is show and market specific, this makes it ideal for accurately gauging on-the-ground performance relative to key competitors or the previous season’s release. Pre-release demand for titles is available daily up to 6 months before the premiere of a show.

The daily nature of demand also allows adjustment in real-time to leverage factors that prove to be extra successful. For example, are audiences reacting especially strongly to the airing of Episode 4? Fuel this with additional marketing material around the key moments that caught interest – behind the scenes/making of, actor reactions etc.

Marketing professionals with a worldwide focus may also want to bring in the momentum metric to their campaign measurement for a global ‘second opinion’. The momentum metric measures the cumulative growth of demand for titles that have been trending upwards in demand for a set period, indexed to the average momentum of the genre. This is a way to quantitatively assess how your title is performing compared to the global competition. Outpacing the momentum of an existing title with a new debut is a good sign your marketing is resonating with global audiences!

Real-time campaign tracking is not defined by a workflow. All the data required for daily campaign performance tracking is easily and simply accessible within an Analyse- level DEMAND360 subscription.

Quantitative post-campaign attribution with the Post Campaign Analysis workflow

Whether as client or agency, accurately assessing the success of a campaign once it is finished is important to improving how campaigns are run in the future. Older attribution methods tend to be last-touch, allocating 100% attribution credit to the last channel a consumer interacted with. This can overemphasise certain channels while effectively discarding the contribution of others. Multi-touch attribution eliminates this by assigning attribution credit throughout the consumer journey.

By combining campaign information and demand data with econometric modelling, we can precisely and quantitively determine the impact of the channels used a campaign. This allows us to uncover the optimal marketing mix for future campaigns.

This is post campaign analysis for the attention economy:

This is the Post-Campaign Analysis workflow. After this workflow, you will have clarity on which channels are EFFICIENT, which are EFFECTIVE and the best MARKETING MIX to deploy for future shows. These are the steps we will take to reach this:

1. We will take the campaign spend and activity metrics.

2. We will combine these metrics with demand in an econometric model.

3. From the model results, we will discover how much demand each channel drove.

4. We will analyse each channel’s spend and activity against demand to determine their efficiency and effectiveness.

5. Considering the headroom of each channel, we will use the effectiveness and efficiency to find the optimal marketing mix for future campaigns.

Econometric modelling of channel contribution

The econometric modelling methodology is by no means new to marketing analytics. However, it becomes exceptionally powerful when the billion+ daily consumer activity data points that comprise demand are used as the model variable. The model allows us to quantify the demand impact of each marketing channel while controlling for outside factors. As the model is custom-built for TV content marketing, this allows an industry-best multi-touch attribution and quantification of marketing impact on audiences.

This equation outlines the demand econometric model:

In the model, we define the daily demand for the show (Yt) as being made up of both the combined effect of the show’s inherent qualities (a1) and the share and volume of demand that each marketing channel drove.

The inherent qualities a1 can be thought of as the show’s ‘baseline’ demand. If no marketing activity was conducted on behalf of the series, then you would still expect it to achieve this level of demand. This is influenced by factors such as the fame of the cast, if it is an adaptation of an existing IP and so on.

We give the model the inputs for each channel (X2t, X3t, etc) – this is either the channel spend or an activity measure such as daily impressions. From these inputs and the demand, the model finds the best fit for the data and yields the constant a1, the error et and the channel coefficients b2, b3 etc.

We are most interested in the channel coefficients. These are each channel’s contribution to the show’s above-baseline demand.

This graph shows the overall demand driven by each channel over the course of the launch campaign for an example series. We see that the TV campaign easily drove the most demand, followed by OOH and Digital while the VOD and Social campaigns drove the least demand. The demand driven by each channel is also available on a daily basis for deeper analysis of campaign flashpoints.

Although demand is a digital metric it does not favor digital channels. As noted earlier demand is comprised of multiple consumer touchstones and favors the biggest consumer point of contact, which in this example is TV.

Now that we know the share and volume of demand that each channel drove, we can determine how efficiently and how effectively each channel performed.

Individual channel effectiveness

To get the effectiveness metrics, we can simply divide each channel’s volume of demand driven by the share of the campaign’s impressions achieved. TO illustrate this, let’s say we have Channel A and Channel B that both drove an identical increase in demand. However, Channel A did this with 50% less impressions than Channel B, making Channel A twice as effective as Channel B.

For our example campaign, the effectiveness of each channel looks like this:

A higher index means that the channel is more effective. As the data is indexed, an effectiveness of 100 is average.

The most effective channel was VOD. VOD drove the most demand for each impression. Compared to the other channels used in the campaign, it was around 3 times more effective. After VOD, the OOH channel was the next most effective.

At the other end, the TV channel actually had the highest number of impressions: 60% of the entire campaign’s impressions were from this channel. But these impressions did not drive a correspondingly high amount of demand, so consequently TV was below average in effectiveness. By far the least effective channel was Social, which was only around 1/10th as effective as VOD.

However, effectiveness is only half of the puzzle. While the OOH channel was effective, it was also one of the more expensive channels, accounting for 33% of the overall marketing budget. As marketing channels have different cost per thousand impressions (CPM) a channel can be effective but inefficient due to high cost.

Individual channel efficiency

Like effectiveness, we can determine each channel’s efficiency very simply. In this case, we divide the share of campaign spend by the demand driven for each channel as shown in this chart:

This data is also indexed, with an effectiveness of 100 being average. Opposite to effectiveness, a higher Efficiency Index means the channel is less efficient.

A high efficiency means that each dollar invested in the channel drove lots of demand, while low efficiency means that each dollar invested in the channel did not drive much demand.

This shows that although OOH was effective, the demand it drove for the cost of the channel means it had low efficiency. OOH has the second highest Efficiency Index, around 80% as efficient as the campaign average. However, a high cost does not mean that a channel is automatically inefficient. The TV portion of the campaign had an even higher spend at 40% of the campaign yet had above average efficiency.

The most efficient channel was VOD which is not only the most effective channel, but also the most efficient. The least efficient channel is also consistent, as Social was the lowest-performing channel for effectiveness too.

Each campaign should balance effective and efficient channels. The next part of the analysis allows marketers to understand which channel ratios are needed for that balance

Balancing overall campaign effectiveness & efficiency

This graph gives the percentage of campaign Spend, Demand and Impressions for each channel over the entire campaign. It also shows the effectiveness & efficiency indexes for each channel from the previous two sections.

As before, a higher number is better for Effectiveness, while a lower number is better for Efficiency.

VOD was the clear winner in both metrics. However, the VOD channel was also one of the smallest channels in play, accounting for 6.7% of the demand driven by all channels and for 4% of the campaign spend. Similar campaigns in the future should invest more resources in this channel. The Digital channel also overperformed on both metrics and so is another good candidate for increased investment in the next campaign.

The comparison also allows more fine-tuned responses. Although OOH was not efficient, if higher campaign effectiveness is required then increased OOH investment can be considered. Similarly, investment in TV could be increased if efficiency is the goal.

Finally, we see that in this campaign, Social was neither efficient nor effective. Unless this was clearly the cause of an unusual external factor during the campaign, in future Social should be deprioritized as a channel for similar show launches.

Optimal channel mix recommendations

We now know which of our channels perform well in each metric, but we cannot simply move 100% of our spend into the ‘best’ channel. Expanding a channel beyond its headroom could quickly lead to diminishing returns, reducing the overall campaign effectiveness and/or efficiency. In our example, the small size of the VOD channel in spend and demand share points to a small headroom, so the bulk of increased investments might fall on the less overperforming but larger Digital channel.

By taking the efficiency and effectiveness results above and accounting for headroom, we can determine an ‘ideal’ campaign marketing mix that gives optimum impact for similar show campaigns in the future.

We can quantitatively determine the market headroom for the channels in our campaign using non-linear saturation curves (s-curves). Hundreds of s-curves are generated by applying transformations to the completed econometric model. Of these, the s-curves with the best model fit illustrate the headroom of each channel.

This chart is an example s-curve for one channel, showing the relationship between an increase in spend and the response in driven impressions:

At low levels of spend, the channel is not driving many impressions. However, there is a breakthrough point around USD30,000 after which the s-curve becomes steep. This is the optimal area, in which each additional dollar invested drives a large increase in impressions. After around USD80,000 the channel becomes saturated and each additional dollar starts to have a strongly diminished impact on the number of impressions it drives.

With s-curves for all channels, we can fully optimize overall marketing spend. For example, perhaps the reason our Social campaign performed so poorly is that it never rose above the breakthrough point. It could be the case that a minor increase in spend would drastically increase the effect of the channel by moving it into the optimal area.

We may also find that some channels are over-saturated. Knowing this, we can reduce their spend to the optimal part of the curve. This retains almost all of impact from that channel while allowing increased investment in a different channel.

The saturation curves, combined with the effectiveness and efficiency of each channel, allows planning of a marketing mix for future campaigns that is optimal. Econometric modelling with the unparalleled consumer Demand dataset ensures that the resources invested into each channel of future campaigns are will yield close to maximum returns.

After this workflow, you will have clarity on which channels are EFFICIENT, which are EFFECTIVE and the best MARKETING MIX to deploy for future shows.

1) Take your campaign spend and activity metrics.

After the campaign for the show launch concluded, we know the impressions and spend for each channel employed.

2) Combine these metrics with demand in an econometric model.

Parrot Analytics modelled the campaign, discovering how much demand was driven each day above what the show would achieve with no marketing.

3) From the model results, discover how much demand each channel drove.

The model showed how much demand was driven on a per channel basis both daily and overall.

4) Analyse each channel’s spend and activity against demand to determine their efficiency and effectiveness.

For each channel, efficiency measures how impactful each dollar spent was, effectiveness measures how impactful each impression was.

5) Considering the headroom of each channel, use the effectiveness and efficiency to find the optimal marketing mix for future campaigns.

With this data, we can adjust the marketing mix for future campaigns, increasing the investment into successful channels without saturating them.

Use the tools to successfully stand out in the Attention Economy

When the marketing mix learnings of the Post-Campaign Analysis workflow are combined with the precise audience segment targeting of the Marketing Planning workflow and the real-time Campaign Tracking unlocked by daily demand metrics, you have an immensely powerful set of tools offering a new approach to content marketing tailored to the audiences of the attention economy.

Discover more about how your content marketing can thrive in 2020 at parrotanalytics.com.