Leiki Newsletter for Digital Media Professionals
Issue 2008/1
Ad Targeting - the Future of Online Advertising
In this issue:
< Previous issue (Solving Online Media's Biggest Challenge) |
E-mail us to subscribe
Rain is falling in the quiet Helsinki harbour on a bleak Thursday afternoon. The offices of engagement agency Toinen are right on the pier, occupied by luxury yachts in sunnier times. The funky and colourful Toinen office is buzzing with productivity and the founding partner Teemu is running through his ten or so daily meetings as we sit down for an interview.
Petrus: What are the biggest challenges for the advertisers regarding the web as a medium?
Teemu: There are two main challenges. For one, the advertisers are searching for the best way for themselves to be visible on the web and how they can get the best results. Traditional banner-advertising is not seen as the best approach as the success of a banner ad campaign in most cases is measured by a number of actions, for example, clicks. Typically the results with banner ads are quite slim, e.g. 99,5% of the users do not react in any way to the banner.
The other challenge is the context: what is the right context for a particular advertiser to be visible? What content would support the message or the products the best? The content on web is mostly  light and quite spread out or even random, and it's hard to find the content and the context that would be the best match to the advertiser's product or message. That's why most advertisers currently have to focus on volume - for example, appear on as
many front pages as possible.
Due to these problems, such as low click-through rates, and problems in the format of the ads - the advertisers feel the people do not pay any attention to banners - the advertisers do not value online advertising that highly. What is strongly needed is a better integration of the content in the media and the advertiser's message. This is why I think technologies like Leiki will become critical as it can combine the editorial content and the advertiser's content and take advantage of both. This is a good way to bring the media and the advertiser together.
How about the challenges on the online media sales - are they the same as what the advertisers face? Does balancing between the editorial team's wishes to perhaps have less ads and the media sales team's need to sell as much ad space as possible cause friction? Are there other challenges or development topics from the media's perspective?
 The biggest bottle neck in online media sales is that online media sales and usage has become too scientific. While media sales and the buyers, advertisers, would usually look for a certain type of exposure or coverage according to their specific marketing or sales strategy, in online media this are somehow forgotten. Instead the online media sales and the buyers turn to numbers of impressions or clicks. This confuses the advertisers - they don't know how the clicks work in their strategy or what can be achieved with them. The online media sales of course have to deal with the pressure of getting their inventory or ad space sold by any means necessary. In problems like these people should just stop and get back to the basics - sell what the buyers need. Don't sell impressions or clicks but sell solutions.
You mentioned Leiki's technology - we at Leiki have launched a new advertising model 'Catalogs to Content', meaning contextually targeting products from catalogs in the online media. How do you see this working and what are the benefits to the advertiser or the media?
 A majority of the advertisers are familiar with the web and know that their customers are using it. As I explained, the problem is though that the advertisers do not know where or how they should be
visible on the web. An advertiser that could place their products to just where they're needed would be the best case scenario - to appear only when their products are relevant. That would remove a lot of guesswork from the equation and it's a very appealing idea to the advertiser.
 By using automatic catalog linking the advertiser always appears when someone is looking for them. This also offers a long-term solution to the advertiser and that's one of the major advantages. With this kind of technology the advertiser doesn't have to worry about finding the right context for their products when the technology does it automatically. The advertiser can give their entire product catalogue to Leiki's system and the system takes care of the rest - each product or product group is advertised in the most relevant context. Media would certainly welcome having long-term partnerships with the advertisers and this type of solution enables that.
We have used our technology to put text links of companies from the Inoa company directory to the end of the articles on Talentum news sites. I feel that it's perceived more as providing valuable, useful information to the readers - not perceived as pushy at all.
I'd agree with that. For example, if I look at a travel article to let's say Bermuda, if I like the article and get interested about the topic I would also certainly like to know how can I get there, how much would it cost or just that I'd like to get more information about it. Why couldn't there be an ad about flights to Bermuda? I think it's only serving the reader.
I'm sure there's a lot concern in the editorial teams about this, that advertisements would compromise the article's or the journalist's integrity. It's a sensitive topic.
Surely that's the case although I think there is no need to be concerned about it as the advertisement is marked as branded content. If the information, the advertisement, is not on the news portal, then the reader has to turn to the search engine and in that case the media loses the reader.
I gather that you see the contextual partnerships between the media and the advertisers as a rising trend?
 Yes - actually I think that media should be more open to partner
with the advertisers on the content front and together create content that brings the best of the both. The bolder media companies have already started to build new portals where the content is developed in partnership with the advertisers. I'm sure this will happen more and more in the future and I think it can benefit the consumers as well.
What is your take on the pricing models in online advertising? Currently I think most companies use CPM-models for display ads and CPC for search ads. Will the pricing models evolve?
 I have quite a strong opinion about this. There are two types of advertising models that are on the web: Firstly there is the advertising that aims to gain and uphold visibility, to bring awareness. Secondly there is transaction-based advertising where the goal is just to invoke some type of action - here CPC, and increasingly in the future also CPA, pricing is applied. I believe that the search advertising will move from CPC to CPA pricing. We are always going to see some hybrid models between CPA, CPC and visibility, depending on the advertiser's goals. But the two models will always remain as separate models - the first one for visibility and brand recognition and the second one for getting results - and in the future there might be even different locations or media for each. All-in-all, I think the results-focused advertising will increase and take over.
Do you think there will be that type of hybrid models for pricing e.g. text links that the display of the link will have one price and clicking the link will have another, a higher price?
Sure I see that happening. The media should always make sure that the display will have a price and will maintain its value. Petrus: How do you see the targeting relates to the costs? I see that the whole idea with targeting online ads is that the cost per action becomes cheaper or even if the costs are the same then the action has more value to the advertiser.
 The web has previously been sold as a bulk media, with the exception of search advertising. The targeting has to be valuable in order for it to make sense to the advertiser. I think automatic catalog linking is valuable as then the advertiser gets visibility next to relevant content, and it will have a direct impact on sales.
I understand targeting in today's advertising is mostly based on demographics, which comes from the old media, especially print, where they do a lot of studies on reader demographics. Another method of targeting is location based targeting. However it's nowadays quite rare that people belonging to a certain demographic or location are all alike. For example, I don't think that people that live in my building are like me or interested in the same things. How does the demographics relate to online media?
The traditional media does use a lot of demographics information and they've had plenty of time to collect the information. The web has not been studied in that manner, at least not that intensively and the web doesn't have a lot of historical data. Also, I think that the web users are harder to put into boxes in that way - a person can look at one web page once, e.g. read one news item from one site, and not return back to the same site for a long time. The web should be a place that would not follow demographics but follow behaviour, such as following how people respond to an ad. The advertisers  could gain a lot more valuable insight from the behavioural perspective. For instance, in a new product launch the advertiser can track on the web just what aspect of the product the users view the most and then use that information to highlight those aspects. I think the web should be used for learning what the users are interested in and the advertisers could build web sites that would allow that. Behavioural targeting however is still not used that much.
Leiki's 'Catalogs to Content' -advertising model is based on contextual targeting of ads. In addition to that, there is behavioural targeting based on the individual's behaviour and social or collaborative targeting based on a group's interests. These are broad concepts and cover various different approaches. Do you have a perspective on these and how they will appear in the advertising mix in the future?
I wouldn't compare the different models as such. Contextual targeting is very useful when you can straight away serve an ad based on what the user is already looking at. Behavioural targeting is also very powerful if you can anticipate what the user is interested in, based on the past behaviour. The problem is that it commonly requires a large volume, which is not always available and takes a lot of time accumulate. Another thing is that usually what can be concluded is largely generic information about a group's interests and not very personal.
We at Leiki have been working on behavioural targeting for a long time and have in fact personal targeting that works very accurately for a single individual user. We can see what topics each individual is interested in based on their behaviour and can serve information or advertisements to them based on their personal interests.
 Well, that is every advertiser's dream! Great advertising is just that, when you can serve the ad as a piece of content, something that the user is looking for.
In early 2008 Fonecta integrated its Inoa company directory, containing approx. 400.000 company titles in Finland, with the Leiki Focus content discovery engine to launch Leiki's new advertising model. This system is the first in the market to enable automatic contextual linking of products from catalogs next to any relevant online content. The system allows any product catalogs (such as e-commerce sites' product lists, "yellow pages" directories etc.) to target ads of each of the products in the catalog to relevant content on other sites, e.g. news or blog sites.
By integrating the Inoa directory with Leiki Focus, the companies in the Inoa directory are advertised by linking the companies with other online content, offering the users a direct access to directory. The targeted Inoa company advertisements are placed next to news articles on various Talentum business and technology news sites such as www.talouselama.fi and www.tietoviikko.fi. This way the most relevant companies in the Inoa directory are automatically shown next to each news article on the Talentum sites. Inoa's company directory is good match with Talentum's news sites as the various news sites cover a wide variety of fields of business and technology, as does the Inoa directory. This allows a comprehensive utilization of the directory.
"Fonecta's B-to-B services, such as Inoa and Fonecta Finder offer a wide range of content from contact and financial information to the actual people in charge. It makes us happy that Leiki has boldly introduced new opportunities to move Finnish web users with us, Talentum and Toinen. As a new feature this enables more detailed targeting of products and services, and certainly a new way for the advertisers to do business", states Jukka Tukiainen, Senior Manager, Strategic Partnerships of Fonecta.
 How does it work?
The new advertising model is based on Leiki's very detailed ontological analysis of content topics and their relevance. This advertising model is particularly helpful for e-commerce product catalogs, any type of directories or content databases. So far, online advertising has either been untargeted banner-type brand promotion or ads targeted using manually defined keywords. This means that online retailers have marketed their services mainly as generic, untargeted brand promotion.
However, with Leiki Focus it is possible to advertise actual products from the catalog or directory on any media site by displaying the most relevant products next to the most relevant content, automatically showing personally relevant items to consumers. The detailed ontology spots the connections beyond simple keywords and allows the catalog owners fully utilise their entire product catalog, without being dependent on keyword matches or 'most popular items' lists.
Taking retail marketing to the next level
 With the 'Catalogs to Content' model the eCommerce retailers' online marketing becomes much more effective than the existing brand promotion, with targeted product advertising seen more as a useful service than just another pushy banner campaign. Showing eCommerce products on media sites instead of just the online store makes a huge difference both for the retailer and the consumer.
Click here to download the case study [pdf]
We're pleased to announce the release of Leiki Focus 5.11.0. In this version, we have put most effort into improving the campaign management features as the software is increasingly used to manage advertising campaigns as a stand-alone product. Also, once again, we have been able to squeeze even more performance out of our already tuned engine.
Key improvements in this version include:
- Improved campaign management support with tools for creating campaigns and tracking their performance
- Statistics such as click-through-rate can now be tracked for individual items, advertising campaigns or ad slots
- Large performance improvements due to a new matching heuristics that significantly reduces the number of operations performed in generating recommendations
- Faster content analysis due to a new pattern matching algorithm
- Standard API extensions such as ability to filter result lists for a variety of customer use cases
Latest Leiki News
New Leiki demo and presentation
Leiki has released a new public demo! See Leiki Focus in action and test how our unique technology intelligently analyses content and finds similar items. Click here to test the demo!
A new presentation, which describes how our contextual and behavioural recommendation engine Leiki Focus works, is also available for download. Download now! (pdf)
Click here for more latest Leiki news
Leiki Online Advertising Survey
Click here to take part in the Leiki Online Advertising Survey. We'd like to hear your take on online advertising and your opinion on the Leiki Newsletter. It will take approximately 5 minutes to complete the survey.
For more information or to subscribe to the Leiki Newsletter please e-mail us.
Leiki Newsletter for Digital Media Professionals is a newsletter about Digital Media and Learning Personalisation published by Leiki Ltd. Leiki provides software technology for digital media to automatically personalise content and community services. Established in 2000, Leiki is the winner of the 2007 European ICT Prize.
|
