What kind of expertise you should demand when buying online advertisement
Text: Sini Kervinen – Aller Media
This blog is a part of SCALED’17 Future Digital Sales Forum guest blog series. Pre ticket sale ends 2.3!
When talking about online advertisement it is good to remember that the end consumer who would potentially see the targeted ad might as well be you or me. “What kind of advertisement would I like to see” and “what prompts me to buy?” are great questions when planning an online campaign. Doing a successful campaign means hard work and testing – no more assumptions about those who, in our opinion, might buy. Let us always remember that no matter how amazing the targeting technology we are using is, there is always a real person at the end of that funnel. If we let technology blind our way of doing online advertisement, the short-term positive results will eventually turn against us. There must always be a professional handling the data and managing technology options; only data and machine on their own will not be able to create the desired outcome without the human touch.
To be able to understand the current state of online advertisement better we need to rewind back in history. Around 8 years ago I worked in a company where I was responsible for Finnish affiliate cooperation. One of my most important jobs was to manually optimize the best fitting ads for selected media sites. I had hundreds of options to choose with multiple prizing models. I had to trust my instincts and reports that were slow to pull and update. I had to look critically at how well the ad messages related to the subject of chosen site were, how tempting the offer was, and what the advertiser was willing to pay for impressions, clicks, actions, leads, or some combination of them. In those days I saw many success stories of advertisers who started small but grew big thanks to excellent ads and hard work along with manual optimization.
Nowadays online advertisement has developed to a very different level. Optimization based on where the ad performs best happens automatically via ad management system, and there is no longer the need for time-consuming manual optimization, as I had described. Ad requests are flying in hundredths of a second finding exactly the kinds of people we wish to reach via the wanted device and in an optimal time of day.
Around 70% of Finnish media sites are premium ads. Programmatic buying was 14% in the first half of 2015; the rest of the inventory normally goes to house ads. Programmatic share is growing rapidly, and I believe that the actual amount of programmatic ads in many media sites is already bigger than the official number. Programmatic buying is a fast, cost-efficient, and easy way to reach the desired target group – so it is not surprising that it is growing in popularity so fast.
The huge growth of mobile usage has changed the way people are consuming media content. More than 40% of the actual page views are on mobile, but the skills to monetise mobile inventory have not been developed as fast as the actual usage. In mobile, the advertised content is able to reach the consumer in very personal stages: mobile is something we look at first in the morning and last in the evening. I personally even watch TV series from my iPhone screen in my bed often before I fall asleep. It seems that the small size of the mobile screen is not a problem anymore; in advertisement, it can even be beneficial when one really has to think about the main message and how it could be visually clear. Even though 50% of the selected target group might stay unreachable, if the advertiser selects to run their online campaign only on desktop.
Targeting and target groups have been one of the hottest topics in online advertisement for some time now. Traditionally target groups are created based on media site categories such as sports, fashion or travelling. Some technologies are doing categorisation automatically based on those sites or articles the consumer reads.
The problem with these traditionally made target groups is that they only scratch the surface of the actual interest. What does it actually mean to know that the consumer is a 30-year-old female from a certain part of Finland and is interested in sports? Where are all the tones of this person’s interest? As an example, we have discussed some time ago with my colleague of the same age, who also is a mother of a young kid, about where we would like to travel next. I wanted to be a mother who travels with her child, so a logical option for me would be Malaga, Spain. My colleague with same demographic background said instead that she would like to travel to London and have a great party. So, if a travel advertiser who had chosen their campaign target group based on assumption where this demography group would like to travel, at least one of us would have most certainly gone badly wrong. The good news is that with modern technology we are now able to reach different areas of interest, no matter of what age or gender the consumer is.
Artificial intelligence is fascinating and the subject of many books and movies. How smart can a machine actually get and what can it learn? I studied computing around 10 years ago. My classmates and I were sitting in a class side-by-side and talking via chat room. Already then, a group of our school students had coded a bot that actually talked with us in that chat room with the nickname “Satan”. Satan was a piece of code that behaved like one of us and answered when questioned by name. We could teach him suitable answers by repeating the same action models enough times. Nowadays only imagination is the limit to what people are able to do with a piece of code.
Nowadays we are using technologies that understand written text like humans do. They understand if the written tone is positive, negative or neutral. If the text is about sports and Beckham, technology understands that it is about football. Sarcasm is something that technology cannot understand yet, but neither can many of us. The best-known technologies are based on ontology, which is a method for studying relations between different concepts, and semantic analysis that studies the meaning of words and thoughts. In the Finnish language, for example, there are about 1 million words all having their semantic meanings. With these kinds of technologies it is possible to get an extremely deep understanding of consumer behaviour and to build unique and real-time target groups that are no longer based on merely assumptions.
Let us think now for a moment about how much technologies have developed in the past years. About 7 years ago in a CeBIT exhibition in Hannover, I had written a text with my eyes, watched halogen television, and a giant computer had scanned me finding out my gender and mood. In this light, the idea that online advertisement and target groups could still be on the same level as then starts to feel absurd. Still, stagnation has reached this business area in some level, which is sad, since online advertisement has a huge impact on the economy globally.
I have been thinking why online targeting is still so old-fashioned in so many ways. Could it be laziness, since earlier it has somewhat been easier when the options where narrower? The advertiser could have chosen sports sections as a target group, made a creative, and started advertising. Once the campaigns ended, he could have known whether it was successful or not. Nowadays, when modern targeting technology gives us the opportunity to make unique target groups for every single campaign or even test what kind of consumers are actually attracted by the product in the creative, the old level of doing this are just not enough anymore. All advertisement in the end aims to increase sales and advanced advertisers nowadays optimize their premium online campaigns based on actual sales also during the campaign; they do not settle just to wait.
Creative has a huge impact on how well the campaign performs. When a selected target group and campaign creative are not matching in the way you hoped they would, you should not hope to get good results either. Still, we blame the targeting technology too often for the results that were not as good as we hoped they would. The popularity of video usage will increase dramatically in the near future because of its power to affect consumer emotions and build brand awareness.
Big data and the way we are able to use it in creating and finding the perfect target group for each campaign will still develop. Combining online and offline data will give us in the near future even more advanced options than ever before. When we, for example, know the address, family size, and occupation, and add these offline facts to an online behaviour such as interest to organic food or desire about buying new Toyota, we are able to bring our understanding of the target groups to a completely deeper level. This does not only give us realistic knowledge about consumers as human beings, but also information on how they behave on different devices and how their interests differ in each.
Big and reliable companies will be playing a big role in the future. Besides the fact that data is used for better online targeting, it is also used for leading, analysing, and product development. The amount of data providers will grow, but the actual jackpot will go to those who have the knowledge on how to deliver the data in the desired form with high-quality combinations. Small data rivers will dry or they will lead to bigger lakes. This new world full of big data opportunities is open-minded; it blurs the line between competitors and potential partners. The more accurate these targeting options go and the more personal the information being handled gets, the greater will be the responsibility for publishers, advertisers, and all middlemen in this area. When this responsibility is carried with honour and the advertisement is done respectfully and in a smart way, then we are heading towards the great future of truly beneficial online advertisement.
About the author
Sini Kervinen, Head of Online Advertizement at Aller Media
Sini’s expertise comes through wide range of experience in online advertising solutions, ad management systems, and different media environments. Last 7 years she has focused on developing online technologies and data targeting capabilities. Currently Sini is working as a Head of Online advertizement in Aller Media and is a Board member at IAB Finland.