| Targeting. It's quite possibly the most powerful | | | | visitors with advertising geared to their interests. |
| factor in any advertising or marketing campaign, and | | | | What's more, if you really want to optimize your |
| for good reason: it works. But to reap results, you've | | | | earning opportunities, you'll carry out contextual |
| got to get it right. And that, of course, is easier said | | | | research for your online content, too. |
| than done. | | | | Let's say, for example, that you're running a food |
| In theory, effective targeting shouldn't be quite such | | | | and wine site and want to write a new article on |
| a challenge. After all, the idea is straightforward: find | | | | wine. You happen to know a lot about German |
| the right product, find the right user, introduce the | | | | whites, but by using a keyword tool to check the |
| two. But as most publishers know, getting those | | | | term 'white wines', you'll notice that 'California white |
| clicks, leads or purchases can be tough. It's estimated | | | | wine' is currently the highest-paying keyword phrase |
| that just a couple of users in every hundred will even | | | | - much more so than its German equivalent. |
| think about going near your ads, which means you're | | | | Which means you've found an ideal topic, since any |
| going to have to work extra hard to persuade them. | | | | contextual ads generated by your content are likely |
| And that's where smargeting comes in. | | | | to be the highest payers. And while you're at it, you |
| Smargeting is targeting with a twist. It's smart, savvy | | | | may as well check out a few other good earners, |
| targeting, but that doesn't mean it's difficult (quite | | | | too. A pretty wise smargeting move, right? |
| the opposite, in fact, which is another reason why it's | | | | Maybe not. Consider the following scenario. |
| smart). | | | | Your visitor - a regular - hasn't yet turned up at your |
| To make a start with a smargeting approach, we'll be | | | | site. They're intending to stop by, but right now |
| thinking carefully about our ads from two points of | | | | they're viewing a blog. It's also about food and wine |
| view: the way they look and the way they perform. | | | | because after all, that's their main interest. |
| There's plenty that's already been written about | | | | There's an ad at the top of the page for Californian |
| ways to enhance ad performance, and any legitimate | | | | white wines, which happens to tie in nicely with the |
| methods that optimize revenue clearly have a place | | | | entry they're reading. Elsewhere in the blog, they |
| in smargeting. | | | | notice an ad for Belgian chocolates. |
| But I'm about to share a smargeting concept that's | | | | They decide to switch to your site, keen to check |
| not as widely used as it should be. | | | | out some recipes. |
| The received wisdom is that 'blending' ads makes | | | | They immediately notice an ad on the homepage. It's |
| them more successful. In other words, if an ad's | | | | the one they've just seen for the Californian wine |
| colors and looks closely match its environment, users | | | | producer. And oddly enough, your feature article is |
| are more likely to click it. The reasoning behind this is | | | | similar to the one they've just read. |
| so-called 'ad blindness', the idea that if ads stand out | | | | No worries - your visitor clicks through to the recipe |
| too much, most users will simply ignore them. | | | | page and starts scanning the latest content. Then |
| However, as audiences become far more used to | | | | they notice an ad for Belgian chocolate. It's in a |
| advertising tactics, it seems they're also learning to | | | | slightly different format, but otherwise the same one |
| ignore blended ads, too - despite our attempts to | | | | they glanced at just a few minutes ago. This is |
| convince them they're part of the content. | | | | almost starting to get a little irritating, and your visitor |
| So perhaps another approach is needed. And what | | | | can't help wondering why they're seeing the same |
| we're going to focus on is which type of ad our | | | | ads over again. |
| users respond to best. | | | | Of course, the answer in this case is because you |
| We'll call this 'psycho-blending'; because the aim, | | | | and your rival publisher were being smart - or so you |
| essentially, is to match the look and feel of ads with | | | | thought - with optimized keyword content. |
| our visitors' personality rather than simply attempt to | | | | Realistically, no two sites are likely to choose exactly |
| conceal them. | | | | the same topics, but even if you're not expressly |
| And why not? Ad campaigns in the offline world are | | | | manipulating keywords, this kind of scenario is |
| very carefully created for specific audiences, with as | | | | growing increasingly common. |
| much emphasis placed on the look of the ad as the | | | | Users surf from site to site according to their |
| bare-bones message itself. Why shouldn't online | | | | interests. And if they visit one or more sites about |
| strategies follow suit? At the very least there's a | | | | the same kind of subject, there's a high chance they'll |
| good chance that certain ad formats are more suited | | | | come across the same ads, too. |
| to your users' overall personality than others. | | | | After all, that's exactly what contextual advertising is |
| Smart testing | | | | about. And even the largest advertising networks |
| The first step is to make sure you know exactly | | | | only have a limited amount of ads to share across a |
| which different ad types are available. There are | | | | rapidly growing number of affiliated sites. All of which |
| many affiliate programs, each with their own | | | | means that, sooner or later, visitors with specific |
| distinctive looks or functionality. Peel-back ads; text | | | | interests will encounter advertising on your site |
| with images; inline text; video. You name it; you're | | | | they've seen many times before. |
| likely to find it. Sign up to a selection of programs | | | | So how do we tackle this? |
| you feel might work - or even some you think might | | | | While waiting for the wider use of behavioral |
| not. | | | | targeting (which evades the issue by serving |
| You'll now need to test these different types. This, | | | | different ads to different users) you're not going to |
| admittedly, is far from an exact science, but using | | | | abandon contextual ads altogether, because they do |
| keywords will help balance results. And despite flaws, | | | | work - just less efficiently as time goes by. |
| even simple testing should tell you far more about | | | | The smargeting answer is to review the idea of |
| smargeting users than you already know. | | | | context and prise it open. To take another look at |
| All you need do is alternate your current advertising | | | | what you're offering and see if you can push things a |
| with the formats you've chosen, then track the | | | | little bit further. |
| results. The implementation is up to you: you might | | | | In other words, it's time to start thinking again about |
| wish to trial new formats on just a few pages, or | | | | your average user profile. What else is likely to make |
| throughout your site if you can do so easily. | | | | them tick? |
| However, you'll want to use the same keyword(s) | | | | In fact, coming up with ideas may be a lot less |
| for each ad type so you can ascertain performance | | | | difficult than you think. Take, for example, our food |
| more efficiently. Ensure that they're fairly standard | | | | and drink site. There's a good possibility that visitors |
| terms so that content is readily available across all | | | | are also interested in tableware; products such as |
| formats; make sure, too, that they're keywords you | | | | dinner services, glasses or cutlery. |
| know attract users - after all, we want to test ad | | | | Perhaps they like to hold dinner parties? If so they |
| types, not the keyword itself. | | | | might also be keen on services such as professional |
| For example, with a term such as 'digital camera' you | | | | flower arrangements for the table. Or specialty |
| should find it simple to set up corresponding ads for a | | | | cleaning products - those red wine stains can be |
| wide variety of formats - product mini-malls, inline | | | | tricky to remove. And then, of course, there's the |
| text, text-and-picture ads, standard banners. | | | | whole area of food preparation itself. Kitchen and |
| Leave your pages long enough to get a handle on | | | | cookery utensils or even white goods such as |
| the stats. Essentially, you're looking at clickthrough | | | | fridges, dishwashers. |
| rate - page impressions divided by number of clicks - | | | | Get the idea? Not only have you found yourself |
| and should be able to track this in your affiliate admin | | | | several potentially valuable niches - segments, almost |
| area or own stats software. You'll also want to allow | | | | - within your food and wine sector, you've also |
| a good number of impressions for each ad type | | | | discovered a way to vary the monotony of your |
| tested; a thousand or so at least, if possible | | | | regular, over-exposed advertising with new and |
| Did you learn anything from these tests? Was there | | | | unexpected messages. |
| a noticeable preference amongst your visitors? And | | | | It's the smargeting way to fight increasing ad |
| more importantly, how did clickthrough rates | | | | blindness - and again, it could be worth investigating |
| compare to the format you've been using so far? | | | | different ad types to showcase such content (a |
| If any ad type produces results that are higher than | | | | mini-mall could be just the thing for those kitchen |
| your current one, you've learned something valuable | | | | products, for instance). |
| about your audience. Either that overall, they prefer a | | | | Don't forget to check how well your strategies are |
| different ad to the one you've been using; or that | | | | working. You might be surprised just how far your |
| changing or mixing formats from time to time | | | | smargeting can take you. |
| provokes renewed curiosity and interest. Either way, | | | | Note on AdSense channels |
| you can use such results to push your smargeting | | | | Admittedly, channels can seem a little perplexing, but |
| further. | | | | they're actually very simple to use. |
| The wider context | | | | For example, if you're planning to position ads top left |
| Most would agree that contextual advertising is a | | | | in your page, create a channel named 'top left' (any |
| mainstay of smargeting. | | | | label will do, but you may as well be specific!) and add |
| In the perfect scenario, you've managed to establish | | | | the new ad unit to it following on-screen instructions. |
| your very own context by promoting a niche that | | | | You'll now be able to check positioning performance |
| visitors actually come to your pages for. | | | | by accessing individual stats for your 'top-left' channel |
| But since that's an exception rather than a rule, | | | | in the Advanced Reports section of the AdSense |
| contextual advertising does, in theory, at least target | | | | publisher area. |