Sabela Media's Lentell talks Net advertising

Sabela Media served its first Web advertisement in September last year. The Australian-based company now delivers more than 100 million ads a month for many high-profile clients (including IDG) and has offices and servers in the US, Canada, Britain and New Zealand. Executive chairman Gour Lentell talks with Russell Brown about trends in Internet advertising, privacy issues and where he thinks the New Zealand Internet advertising industry is at.

Sabela Media served its first Web advertisement in September last year. The Australian-based company now delivers more than 100 million ads a month for many high-profile clients (including IDG) and has offices and servers in the US, Canada, Britain and New Zealand. Executive chairman Gour Lentell talks with Russell Brown about trends in Internet advertising, privacy issues and where he thinks the New Zealand Internet advertising industry is at. I should declare an interest right up front - which is that you're now, of course, serving IDG New Zealand's Web ads as well as those for IDG Australia. It must be an attractive element of your business that you can offer co-ordinated trans-Tasman campaigns. We definitely view Australia-New Zealand as a regional market. We've been active to some degree here, but I think we're about to ramp that up - in particular by hiring someone here. We've been servicing out of Sydney, but that can only go so far. What's you take on the online advertising market here? New Zealand seems to me to be about where Australia was six months to a year ago, in terms of level of activity and awareness. People have been wanting to know when the kick's going to come in. In Australia we definitely experienced a kick this year - 300% growth this year compared to last year, much bigger deals being done, a broader base of advertisers and Websites - everything. I wouldn't say it's mainstream in Australia yet but it's certainly moved up in the consciousness of advertisers. It's inevitable that will happen here. It seems that the research is a lot better in Australia - there's a lot more information available. There are three or four active research companies now in Australia that have dabbled here a bit but haven't made a real commitment. It all comes down volume and whether it's worthwhile from a business point of view. How important is it to what you're doing that there is some independent means of assessing the market? There are two answers to that question. One is from a media buyer or advertiser's perspective - they're used to having certain information and research tools available to them in traditional media and they use them extensively in terms of helping plan their media. They feel very uncomfortable when they don't have something reasonably equivalent for online. They feel like they're flying blind. The issue from that is how on earth you measure the results of a campaign - which is where we come in. I'm not a media person by background, but it seems to me that it's fundamental to their psyche that they need some means to assess the market and help them plan. The thing that's very different about online compared to traditional media is what we can do in terms of delivering and tracking ad campaigns. It's definitely a one-to-one marketing medium. We deliver all the ads to all the users for the campaigns they run - and measure everything. The best measurement they can do in traditional media is based on sampling and surveys and clever statistical analysis to extrapolate the data. Whereas online we measure 100%. I very much think that the more an agency or media buyer measures their advertising that way, the more they'll start building up a significant database of their own, based on 100% of their results from the campaigns they do. That will change the nature of their research. The analogy I use is to imagine if a TV advertiser could book their timeslots but then essentially construct the TV so every single viewer came and fetched the actual ad clip from the advertiser's own video server. Thereby enabling the advertiser to measure exactly and entirely who saw their ads, how many people there were, what kind of profile it was and how they responded to the ad. How much can you really tell agencies about who saw their ad and how they responded? We can certainly measure every response - and not just response to an ad but tracking it through to measure customer conversion. Are they acquired as a customer? We can also profile that audience in interesting ways. One of the hardest things to do is profile them in terms of demographic data, but we can certainly do it by geography, by country, how they responded, what their interest profiles were. We can do a lot of profiling based on behaviour and inferred interest. The fact that it's difficult to measure demographics, is, of course one of the things a traditional media buyer finds unsettling. But you can argue that if you're targeting a campaign at people who are interested in sport for example, that's what you really want to know, rather than a conventional demographic breakdown. That's a trend across media in general, though, isn't it? Demographics are becoming less important, psychographics more so. Yes. And the opportunity to do psychographic profiling on the Internet is significant. Another analogy I'd make is that if you look at the role AC Nielsen plays in traditional media, a lot of auditing and research is based on profiling the audience that consume certain types of media - magazine, TV, whatever. A media buyer places an ad in that publication on the basis that potentially 100% of the audience that consume that publication will have an opportunity to see the ad. Whereas on the Internet, where a site has 10 million or 100 million pages views, what really matters is who sees the 100,000 ad impressions booked by the advertiser. They're not broadcasting the ads to everyone who uses that site. To what extent are you depending on people doing things like clicking on banners? I use the Internet all the time and I don't click on many banners. Obviously, one of the fundamental things we do is measure clicks. The industry regards clickrate as one of the fundamental measurement points. Average clickrates are often quoted at being somewhere around 1% - but we see them vary enormously, from half that to 10 or 15%. We've seen good examples of campaigns that really do work and offer good value to the advertiser in terms of cost per acquisition or whatever it is they want to measure. There are many people that don't click, but a significant number who do. How do you tailor a banner to attract interest like that? There are a number of ways it works. We do ad serving for publishers, managing the entire ad inventory on a Website, and also for ad agencies. In the publisher context the targeting flexibility is much greater because you're selecting every page view or ad request in terms of the best ad to show that person. On the agency side, where we're serving all the ads for a campaign onto any Website where they've booked that media, it's fundamentally that Website that's selecting the user in the first instance. We can then optimise the creative deliver to that user as appropriate. A good example is a bank we're working with in Australia, whose ad spend is significant. They're a big site and they do a lot of things online. A lot of their campaigns are acquisition-oriented, trying to acquire new customers. We can work with the bank, deliver all their campaigns, track the responses, track customer conversion - and also use our profiling capability to start tagging or flagging individual users as acquired customers. So the next time - which might be seconds later - that we see that person in the context of the bank's ongoing campaign, we can select a different creative for that person, reflecting the fact that we know they're a customer. What's the technical means for doing that? It's basically based on a profile built against a cookie. Every time we serve an ad to a user we try an place a cookie on their browser. The only thing in that cookie is a unique number so we can identify that user. We simply don't try and capture any personal data about that user - name, address, telephone numbers, whatever. We never want to be able to identify the user as an individual in the real world. So all the profiling's anonymous, and within that we do some generic profiling which is built up over a period of time. In addition, our customers can put small bits of data into that profile - to flag that user as being a customer of particular services. There's always the chance the user will blow the cookie away or use multiple PCs, but statistically there's a very high percentage of that data that will continue to be relevant. Can you use a cookie placed in one campaign as part of another campaign? Potentially we could, but we don't. We have a very clear policy. Our profiling model is not to collect data from any client and use it to the benefit of any other client - or ourselves. In the bank's case, they would set the data attribute and they would be the only ones able to use it and report on it. There's no way any other client of ours would be aware that value was in that profile. What say there or four clients got together and said, we want to share this information, pool what we know? There are two issues there. If our clients indicated co-operation and wanted to do that, that's the only condition on which we'd even consider it. We'd also have to consider what they're trying to do and whether is compromised any privacy issues. We have a very clear privacy policy and we're audited by Price Waterhouse Coopers, who, amongst other things, review that privacy policy. They make sure we do what we say we do. We're really trying to use the media in ways that are positive for the users as well as the advertisers.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.
Show Comments