Campaign planning - The tool that does it all

Accelerom has developed a new algorithm that calculates the optimal multichannel marketing mix. It works for products, target groups and markets in B2B and B2C, and delivers verifiably better results than best practice, the company promises.

The new 360° Touchpoint algorithm simplifies media and marketing planning quite considerably. It enables companies, organizations and institutions to review their current marketing mix or find out where there is potential for improvement or savings. This means that either more impact can be achieved with the budget used or the same impact can be achieved with a lower budget.

And this is how it works: At the push of a button, the algorithm calculates uniform key figures for millions of combinations and compares their impact - whether for different target groups, product categories or countries. The result is, for example, the most effective media combination for the previously defined target. Through its calculation, the algorithm reveals via which media and other touchpoints online and offline customers are led to purchase, and assesses the combined effect of different online and offline touchpoints both qualitatively and quantitatively. The spectrum of instruments analyzed ranges from owned, paid, and earned media to the classic media mix, online campaigns, POS, measures in sales and CRM, and social media. This total audience measurement is the unique selling point of the algorithm.

Unbenannt-16_31
larger

Better than best practice

In contrast to conventional methods, the entirety of the contact points examined by the algorithm is not limited to advertising media or online activities, but up to 100 marketing and sales measures can be included in the analysis. In this way, the algorithm delivers verifiably better results compared to best practice. Accelerom has already been able to prove this with existing customer projects.

The algorithm's multi-layered simulations are made possible by combining various measurement methods from artificial intelligence, machine learning and metaheuristics. Accelerom developed the algorithm over several years in collaboration with NZZ and the multidisciplinary "Institute for 4-D Technologies (i4Ds)". The development research project was funded by the Commission for Technology and Innovation (CTI) of the Swiss Confederation.

Anne-Friederike Heinrich

Unbenannt-15_34

"The output of the machine is better than best practice".

Werbewoche spoke with Christoph Spengler, CEO of Accelerom, about the background of the new 360° Touchpoint algorithm.

WW: How long did you work on the new algorithm?
Christoph Spengler: We launched the first CTI-funded project in 2011. The main focus was on the topic of total audience in the multi-channel world. There are already almost too many analyses on this, but we wanted to look at it holistically, i.e. bring together online interaction points owned, paid and earned as well as offline in a single currency. In 2013, we won another research project that focused on artificial intelligence. And this ultimately gave rise to the new algorithm. 

What requirement did you start with?
We wanted to develop a tool that compiles the best mix from the entire potpourri of interaction points - ad, billboard, TV spot, website, etc. - as a recommendation in the decision-making process. This requires a huge database from which the best multi-channel mix is calculated with the help of several algorithms, and this across all categories, on- and offline, owned, paid and earned. And after several validations in practice, we can say that the output of the machine is better than best practice.

What information do you have to enter and what information do you get? Can you demonstrate this with an example?
Gladly: You want to develop a multi-channel mix, a strategy or a campaign with the target group of young people in the beverage sector that generates attention, promotes engagement and boosts sales. We have about 100 touchpoints to choose from, spread across advertising media, POS, social media, etc. Furthermore, we still choose that we don't want a mix with 30 instruments, but one with a minimum of 10 and a maximum of 15 instruments. After this input, the machine starts to calculate. The output is the mix with the 13 most effective media for attention, the best social media combination for engagement, and the mix of measures for sales at the POS. The tools that the algorithm identifies are the ones that are most effective for achieving the defined goals, not just for generating attention, but for carrying the consumer from attention generation to transaction. We call this the top conversion path.

Is there always just one?
No, the output is three top conversion paths with different online and offline interaction options - because target groups do not always behave the same way, of course.

But I still have to decide for myself what measures I take ...?
You're right, things can get a little scary. But of course, the top conversion paths are primarily an orientation aid for us to subsequently develop a campaign or other strategies. That's an extreme help, but it's still a long way from implementation. For that, you need creativity and a lot of other things that a machine can't do. In addition, there is another dimension: the cost. We judge the top three conversion paths by cost, and of course there are cases where our customer says: Top 1 I can't afford at all. But our algorithm makes certain calculations possible in the first place: a customer operates in three countries or markets, has two target groups, six product categories - that quickly becomes confusing. And since everything can be measured qualitatively and quantitatively, the results are also verifiable.

How long does the algorithm take to calculate the top conversion paths? 
Depending on the task, this can take a few minutes or even longer. It depends on how many calculations have to be made.

But how exactly does the algorithm calculate which is the most effective marketing mix?
For each touchpoint, for each instrument, we collect performance values from our customers, for example, reach, attractiveness, performance information. This information is collated in the algorithm, assessed, additional values are added, what does not offer added value is excluded, values are added again, and so on.

How do the results of the new algorithm differ from best practice?
In best practice, of course, you don't have combined values, you can't compare a website with an ad. Because of this information gap, you can't make a clear assessment in multi-channel communication. So in Total Audience, there is a relatively large amount of information missing today.

So the previous planning of the marketing mix went out into the blue?
Yes. Even if I don't make many friends now: It's a patchwork. Everyone is measuring something somewhere. But the question is: Where is the measuring point? Our measuring point is the actual buyer of products, not the TV viewer, not the newspaper reader, and not the Internet surfer. For that reason alone, we have a much higher degree of precision. It's like a racing bike: What kind of gears do you have, what kind of fork do you have, what kind of frame do you have? That ultimately determines how well and how fast you ride.

Sounds like an egg-laying jack-in-the-box ...
At the present time, this statement can be left as it is. Of course, the algorithm does not answer all questions, because we are moving at a certain altitude. But it offers the best prerequisites for strategic control of the marketing mix.

What questions remain unanswered?
With certain topics, there are indications where you have to go more in depth. For example, if you ask me why a certain target group is not satisfied with an interaction, then you have to look more closely there. But there you would also reach a limit with other instruments. Often it's enough to draw cross-connections. For example, the contact form is very important for a target group - but the customer doesn't have one. Problem solved.

Have you already been able to verify in practice whether the mix combinations suggested by the algorithm are more effective than classic success methods?
We tested the algorithm on data from old and current customer projects and on strategies that have been implemented. And we came to the conclusion that we get better results this way than with best practice.

How can the algorithm assess measures that have not yet been used, e.g. social media or online sales?
We don't just measure what our customer does, we look at what is being done in the entire target group. And in this way, we also see what the qualitative and quantitative benefits would be if the customer were to implement a measure that he has not yet used. For example, if a customer doesn't have an online store, we can tell them very specifically how many people they would reach if they had one. And we can also tell our customer how best to market his store to this target group.

You claim that you can indicate quite precisely which are the optimal marketing and media measures. Can your customer ONLINE can you be one hundred percent sure that the measures you recommend will have the desired effect?
Yes, our touchpoint analysis has a high degree of precision. In recent years, social media measures have been in vogue. We were now able to use the algorithm to precisely predict what has developed in this area, both positively and negatively. We got tangible information out of it, for example that a target group cannot be activated with certain means used or that engagement will be very weak. So we can make quite precise statements about the potential of measures and channels. There is also the case that customers want to reduce and focus their measures. Then, of course, they can eliminate one or more instruments. We can determine which ones would be best. Impact thresholds and advertising pressure are also becoming more and more of an issue. Some clients therefore have us check their media plan.

How much does the new service cost?
Prices vary depending on the consulting project. However, the costs are roughly comparable with investments for market research. The projects with 360° Touchpoint all have a positive ROI per se, because the various measures can be reduced, bundled and thus more budget remains for the individual measure. Accordingly, their impact improves.

Compared to other countries, is your development state of the art or a revolution from Switzerland for international marketing and media planning?
We have held talks with many protagonists from the media world, including internationally, and all of them have confirmed to us that they have never seen a planning tool like this before. We think we can use it to gain a foothold internationally as well. There is already a lot of interest.

Interview: Anne-Friederike Heinrich

printausgabe_6

More articles on the topic