heisel.org

Beyond the click-through

Wednesday | November 5, 2003 | 3:07 pm  

The New York Times has a good arti­cle about the ever chang­ing face of adver­tis­ing met­rics.

Google is now offer­ing a piece of code that adver­tis­ers can place on their site to track users that come from Google-​placed ads.

The arti­cle talks about how this is lead­ing to more effec­tive campaigns.

True, the data it pro­vides is a very valu­able metric. But I’d cau­tion man­agers not to base all of their deci­sions on this glo­ri­fied click-​through data.

I have two words for ad man­agers focus­ing on click-​through per­for­mance: Trin­ity Homes.

What? You ask.

While I was work­ing at The Indi­anapo­lis Star, Trin­ity Homes had pur­chased the top right ad avail on our home page.

The ad was a square, non-​moving, non-​flash, non-​jumping, non-​beeping, non-​everthing ad. It was a purple square with the company’s tri­an­gle logo and their name.

That’s the power of brand­ing. I saw that ad, I still remem­ber that ad, and prob­a­bly will for years to come. And, I assure you, if I’m ever back in Indi­anapo­lis and look­ing for home, I guar­an­tee I’m going to look at Trin­ity Homes’ Web site.

This is a round-​about way of coming to my point: there is more value in online adver­tis­ing than the click-​through alone.

There are four key things that have to be in place for a poten­tial buyer to act on an ad and make a purchase.

The first three we can borrow from one of my favorite shows: motive, means and opportunity.

  1. Motive – Not to kill, but to buy. As a poten­tial pur­chaser I must have a need or a want for a type of prod­uct. When I see adver­tise­ments for the best com­puter, ever I real­ize that I want a new G5.
  2. Means – I must have the means to pur­chase a prod­uct. In the case of the G5, I don’t have the money :-(.
  3. Oppor­tu­nity – This is often less of a factor online, but worth men­tion­ing. Can I get to your store? Is your site down? Do I have the time to go there (either online or bricks and mortar)? Will you sell to me (ask about British folks who would like to try the iTunes Music Store)?

If an ad can reach a poten­tial buyer who meets all three cri­te­ria — a daunt­ing task, when you think about — there’s the fourth and final one to con­sider — product.

4. Prod­uct – Does the prod­uct have the fea­tures, price, qual­ity, etc. that the buyer wants? Meet­ing this is the onus of the adver­tiser, it’s com­pletely out of the con­trol of the pub­li­ca­tion car­ry­ing the ad. Let me put it this way, no matter how many ads Microsoft puts on TV, in print, and on the Web. I will never, ever, ever buy Long­horn. It’s simply not an OS that meets/will meet my standards….

Now pre­vi­ously, click-​through rates could mea­sure how many cus­tomers that met the first three cri­te­ria were coming to an advertiser’s Web site.

If you had the moti­va­tion to buy a type of prod­uct, the means to buy it, and the oppor­tu­nity — on the Web this mostly means you’re will­ing to surf away from the site you explic­ity set out to view… — then you might click on an ad.

But now, with post-​click track­ing, adver­tis­ers are going to demand that their cam­paigns deliver the impos­si­ble — view­ers who have motive, means, oppor­tu­nity and are com­pletely happy with the prod­uct the com­pany is offering.

This is a sim­pli­fied model, to say the least, it doesn’t account for online fac­tors like a fear of credit card secu­rity, users lack of affin­ity for paying ship­ping costs, etc. It also doesn’t account for whether the ad is well crafted, on mes­sage, etc.

Let’s revisit Trin­ity Homes, shall we.

If I need a house (motive), and I have the money for one (means), and I’m in Indi­anapo­lis (oppor­tu­nity), AND Trin­ity sells homes in a neigh­bor­hood I want to be in that are the right size at the right price and that are built well (prod­uct), then I might buy one.

If I were to buy such a house, and if they were only paying The Star for click-​throughs, then Trin­ity would have gotten free adver­tis­ing, The Star would have been gypped, and Trinity’s met­rics wouldn’t have me on their radar at all.

Dis­claimer: I have no idea what Trinity’s con­tract with The Star is like, this is just an exam­ple.

Because, if I need a house, and I have the money, and I’m in Indy, and they’re good houses, I’m not going to go back to the newspaper’s Web site just to click on the link.

I’m going to do like anyone else would, I’m going to Google Trin­ity Homes, because I will have remem­bered that purple box with the white tri­an­gle logo.

That, my friends, is just the tip of the ice­berg called “branding.”

And that’s why pure click-​through rates/models are wrong…

Brand­ing may be less easy to track than click-​throughs, but there are ways. The eas­i­est is to run a cam­pa­gin and see if your visits go up. Stop the cam­paign. Start it again later, if your traf­fic goes up then — tada — your adver­tis­ing works.

More publisher-​friendly meth­ods include those men­tioned in the arti­cle, like Gamespot’s.

For instance, when Elec­tronic Arts is adver­tis­ing Madden 2004, Gamespot lets the com­pany know how many people are hit­ting cov­er­age about the game as well as giving them traf­fic fig­ures for Madden’s com­petit­ing products.

Permalink | Comments (4) | Categories: Business, Management

Comments

aaron wall November 10th, 2003 | Link to this comment |

On big ticket items your point is well taken. On smaller items though I do not think it is too out of reach to expect many of the things to fall into line. Espe­cialy if the user types in searches like “buy staind cd”.

It must be some­what effec­tive if over 100,000 people are doing it!

Chris Heisel November 11th, 2003 | Link to this comment |

True, price has an influ­ence on the deci­sion to click through an ad.

But at the same time, another strong factor to con­sider is the way people surf. They’re out look­ing for infor­ma­tion — so on a news site, if some­one came look­ing for the latest sports scores, there is a lot of trac­tion work­ing to keep them in the sports section.

Its going to be dif­fi­cult to get them out of the sports sec­tion, let alone to another (adver­tise­ment) site…

That’s where having good context-​senstive ads can help alot, but there’s still a lot to over­come in order to deliver a reader to an advertiser’s site…

Honymann November 12th, 2003 | Link to this comment |

Chris,

the way you pro­pose to track the “branding effect” of a banner cam­paign is still highly inac­cu­rate does not help any­body who is on a really tight budget. Rather on should prob­a­bly work with a bonus (expressend in per­cent) on the clicks received through a compaign.

This bonus prob­a­bly depends on fac­tors such as: - is the brand/product adver­tised well known or not (pos­i­tive effect) - the inten­sity of adver­tis­ing (pos­i­tive effect) - is the adver­tis­ing con­cen­trated on a few sites (pos­i­tive effect etc.

This at least gives you the chance to cal­cu­late a “cost per customer” (even if it is an esti­ma­tion) that helps you to run your mar­ket­ing on a day-​to-​day basis.

Chris Heisel November 13th, 2003 | Link to this comment |

I agree that click-​throughs should be seen as a bonus.

But unfor­tu­nately they seem to be the pri­mary metric for mea­sur­ing suc­cess in the online world.

Rather they should be a bonus, as you said, to the impres­sion. Because I’d argue that the act of merely seeing the ad is a value to the cus­tomer and pub­lish­ers should be rewarded for that as well as click-​throughs.

Disclaimer: I work at The Atlanta Journal-Constitution. The opinions expressed here are my own, and do not reflect those of the AJC, Cox Newspapers, Cox Enterprises nor any other party.

Mail carrier N. Sorenson delivering Christmas mail through the snow. (Chicago Daily News/Chicago Historical Society)