B1BLOG

With search engines char­ging for clicks and not impres­sions and typ­ic­ally being con­sidered a dir­ect response chan­nel, you may be for­given to think that there’s little need to optim­ise for click-through rate (CTR). How­ever, since CTR is the major­ity factor to determ­ine the so-called qual­ity score, which impacts the cost-per-click (CPC), this will affect the suc­cess of the cam­paign; regard­less of if object­ives are sales, leads or traffic.

Why are search engines doing this? The adverts CTR is an import­ant sig­nal of “qual­ity” for search engines. The logic is that if the advert matches the users query, they will click it. There­fore the higher the CTR, the more rel­ev­ant the advert is assumed to be. And of course, since the search engines charge advert­isers per click, their rev­enue increases.

The fol­low­ing chart out­lines the main factors that determ­ine the qual­ity score for a keyword on Google Adwords. This qual­ity score is then used in an algorithm that determ­ines click-cost and pos­i­tion.

Due to this, it’s very pos­sible that an advert appear­ing in top pos­i­tion is pay­ing less per click than the one below. The search engines are of course max­im­ising their rev­en­ues by the higher num­ber of aggreg­ate clicks that the top CTR advert is producing.

To make things more com­plex for advert­isers, con­sider the fol­low­ing scen­ario: an advert with lower CTR is con­vert­ing very strongly on the web­site, as it effect­ively pre-qualifies the traffic. In com­par­ison a high CTR advert for the same cam­paign con­verts poorer, as it doesn’t seek to fil­ter out unwanted traffic at the ad level, but instead on the land­ing page. Which should you run?

At this stage you would need to cal­cu­late the actual cost-per-action and volumes that can be achieved, depend­ing on objective.

In prac­tise, it’s dif­fi­cult to make this type of dir­ect com­par­ison on a live cam­paign, due to the ever-changing nature of com­pet­i­tion, qual­ity score, pos­i­tions and costs. How­ever, from exper­i­ence, typ­ic­ally for b2c the bene­fit of a higher qual­ity score from optim­ising ad cop­ies towards CTR (with res­ult­ing lowered CPC and higher pos­i­tion) out­weighs the altern­at­ive, whereas for B2B it might not be the case.

The Independent’s media sec­tion today has a fea­ture on “Why we don’t make good ads any­more” – you can read the online ver­sion here.

They’ve inter­viewed the usual adland sus­pects (Frank Lowe, Mar­tin Sor­rell, John Hegarty, Tre­vor Beat­tie etc). The premise behind the art­icle is that ads today aren’t as good as they used to be – in the days of the Milk Tray man, Hovis bread deliv­ery boys, Smash and all the oth­ers that make up the pan­theon of golden age greats. The respond­ents are some­what divided. Frank Lowe’s com­ment reads as a lament for times past, the oth­ers tend to cri­ti­cise the lack of time that per­vades the industry and are split on whether online is cre­at­ive enough yet.

These kinds of art­icles are not new. Each gen­er­a­tion of cre­at­ive seems to feel a need to talk up the dumb­ing down of the latest crop of work. More recently there has been a flurry of rants as tra­di­tional adland has seen digital begin to eat its lunch and many cre­at­ive dir­ect­ors find it dif­fi­cult to adjust to a world where many people really, really don’t like TV ads (and where they can almost totally avoid them should they so choose).

There is also the wide­spread per­cep­tion that good work = awards. And, of course, some awards high­light some very good work. But, per­son­ally, I fear that many awards cre­ate a kind of self-fulfilling proph­ecy – estab­lished cre­at­ives award­ing work that they would like to have cre­ated (ie tra­di­tional advert­ising). I remem­ber some years back being on a judging panel where the dis­cus­sion got round to “Is this an (insert awards’ name) kind of ad?” and to what would get the best reac­tion at the awards’ night din­ner. This can’t be healthy.

Of course things are chan­ging (it was ever thus). There has cer­tainty been a shift in the kind of work that gets us most inter­ested these days. It doesn’t tend to be the big ad cam­paigns so much any more. It’s the dif­fi­cult, gnarly prob­lems that need clever, eleg­ant answers. It’s the left-field, asym­met­ric cre­at­ive strategies that refuse to fight on a level play­ing field.
It’s the ideas that are not only media neut­ral (or any of the other buzz terms) but which seek to cre­ate new media and forge new connections.

Time, as many in the Independent’s art­icle point out, is an issue. We don’t have the lux­ury of long, draw-out devel­op­ment any­more. But is this such a bad thing? I don’t know. Yes, I love to see beau­ti­fully craf­ted work that’s had oodles of lov­ing care taken over it. But on bal­ance I’d prefer to beat a com­pet­itor to the punch with some­thing quick, dirty and effect­ive. Bet­ter still, I’d prefer to get in early and set the agenda for a product or cat­egory. All the kern­ing in the world won’t match up to bene­fits like these.

Everything these days is in beta (as Rus­sell Dav­ies com­men­ted some time back). The days of 100% fin­ished, totally locked down cre­at­ive cam­paigns are his­tory. Instead, today’s com­mu­nic­a­tions are messier, trick­ier and more inter­est­ing for it. Cre­at­ives might make fewer good tra­di­tional ads any more but with so many oppor­tun­it­ies to do some­thing bet­ter and faster I can’t say I’m going to lose sleep over it.

Last week, I got the chance to hang out at the Future Mar­ket­ing Sum­mit 2007 in Lon­don. The theme was integ­ra­tion and it man­aged to attract a host of prac­ti­tion­ers from a decent vari­ety of agen­cies – large, small, digital, tra­di­tional etc. While it was clear that most agen­cies haven’t fully cracked integ­ra­tion, there was a con­sensus view that we are all headed in that dir­ec­tion – always good to hear when you work with an integ­rated agency.

I won’t go into a blow by blow report of the sum­mit, you can find one of those here.

A couple of things struck me. One was about the rela­tion­ship of advert­ising to design. For two dis­cip­lines that are so close on one level, on most oth­ers they often appear to come from dif­fer­ent plan­ets. I’ll post more about this soon, I want to mar­shal some thoughts first.

The other came from a com­ment that Rus­sell Dav­ies made. The panel was dis­cuss­ing who was good at integ­ra­tion in action (as opposed to those who simply talk a good game). Crispin, Porter + Bogusky came up as the cur­rent poster child of the integ­rated agency world. Rus­sell made the point that there was no magic for­mula to their suc­cess, that they obvi­ously had a dens­ity of tal­ent but that their real skill was that they got on and actu­ally made the ideas real.

This thought was bril­liantly illus­trated later by Tim Ashton of Anti­dote (who I felt was the closest of any­one at the event to get­ting the whole integ­ra­tion thing right and was the one presenter to make me feel down­right jeal­ous). He presen­ted a case study around We are what we do and the Change the world for a fiver book. It was a great example of just get­ting on and doing it. I’m sure the charity-ish nature of the book helped in their abil­ity to beg and bor­row mater­ial but the atti­tude of just get­ting it done that Tim talked about was inspir­ing. If they needed a shot and couldn’t get it, they took it. Many agen­cies wouldn’t dream of doing this, they’d want it all pol­ished and per­fect. But the res­ult Anti­dote achieved was cer­tainly good enough.

This move against overly pol­ished work, towards work that’s always work in pro­gress is a massively lib­er­at­ing thought. Yes, it’s still got to look good and has to have a great idea behind it but not be so ‘con­struc­ted’ that it feels like spin. Jay Chiat is often quoted as say­ing “good enough isn’t good enough” but today, maybe it is. If the think­ing is right, if we make the right con­nec­tions with the audi­ence, if we’re authen­tic, then we can begin to loosen the grip of the cookie cut­ter brand manual.

I can only see this as a good thing.

TED ranks as prob­ably the greatest show­case of col­lect­ive smart think­ing on the planet. It’s held in Monterey, Cali­for­nia every year and attracts some of the world’s most inter­est­ing thinkers and speak­ers. The 2007 TED is immin­ent and fea­tures the likes of Bill Clin­ton, Edward deBono, Lawrence Lessig, Paul Simon, Phil­ippe Starck, Richard Bran­son – the list goes on and on.

I would love to go but the time, dis­tance and con­sid­er­able expense kind of rule it out. How­ever, as the next best thing, you can see many of the pre­vi­ous years’ present­a­tions here. Warn­ing though: this is a pro­ductiv­ity black hole – I’ve spent some ser­i­ous time in the last week check­ing these out (I’ll put it down to training).

It’s an amaz­ingly diverse and eclectic group of speak­ers. One minute you can see cos­mo­lo­gist Mar­tin Rees talk about the uni­verse, the next see Wired’s Kevin Kelly draw par­al­lels between evol­u­tion and tech­no­logy. The many oth­ers include Richard Dawkins, Mal­com Glad­well, Steven Levitt, Nich­olas Negro­pote and Al Gore. Hope­fully the 2007 present­a­tions will become avail­able soon too.

Of course it puts my inferi­or­ity com­plex into over­drive – but that’s a small price to pay.

There’s an inter­est­ing bit of research in the latest Brit­ish Psy­cho­lo­gical Soci­ety’s bul­letin (always worth a read – RSS feed here). Research­ers have shown that most people’s gen­eral bias towards the left-hand side of space could be dis­tort­ing their responses to surveys.

In a sur­vey ask­ing stu­dents to rate their exper­i­ence of uni­ver­sity, the research­ers used a 5-item Likert scale ran­ging from def­in­itely dis­agree to def­in­itely agree. Except of course they swapped the dir­ec­tion of the scale for half the recip­i­ents (in that sneaky psycho-research kind of way).

The res­ult was that the stu­dents using the scale that star­ted with ‘def­in­itely agree’ chose that as their answer 27% more often than those with the scale begin­ning ‘def­in­itely dis­agree’. This is, by anyone’s stand­ards a pretty sig­ni­fic­ant shift.

I can’t help won­der­ing what implic­a­tions this could have for the way we struc­ture inform­a­tion in advert­ising. It’s inter­est­ing that so much of the tra­di­tional emphasis in ad circles is on the right hand page. Even within a DPS, there is a def­in­ite bias towards the right. Maybe this is all, well, wrong.

Some­thing to think about.

You can read the BPS art­icle on the research here.

Order the full find­ings report here.

In their strategy+business emag this time, Booz, Allen, Hamilton has a report: “Smart Spend­ers: The Global Innov­a­tion 1000.” Ostens­ibly this is a report on why some com­pan­ies (about 10% in the study) get expo­nen­tially bet­ter returns on lower levels of R&D invest­ment. Strategy+business is usu­ally a pretty good source of insight so admit­tedly my hopes were high.

There are some inter­est­ing facts in the piece. The top 1000 R&D spend­ers account for some 85% of R&D spend­ing as a whole. The authors point to the fact that there is vir­tu­ally no cor­rel­a­tion between per­form­ance and R&D budgets (the only one being gross mar­gin – yet the fin­an­cial value of this is sel­dom cap­tured). Many com­pan­ies waste too much time rein­vent­ing the wheel and see ideas get mired down in organ­isa­tional quick­sand. And many simply don’t under­stand what their cus­tom­ers want or how to com­mu­nic­ate with them.

All inter­est­ing stuff. But how about inform­a­tion that com­pan­ies can use to do some­thing about it?

The recom­mend­a­tions appear to boil down to:

• Get good at all parts of innov­a­tion ‘value chain’
• Go for flat structures
• Think about cus­tomer needs
• Be quick

Bey­ond that… well… every­one does it dif­fer­ently really.

This has the feel of a scen­ario where some very expens­ive research has been com­mis­sioned, but which hasn’t answered the key ques­tions. So you get lots of inform­a­tion but pre­cious little insight bey­ond a kind of innov­a­tion 101.

I would want to know about the cor­por­ate cul­tures that breed innov­a­tion. I’d want to know who they recruit, how they get the right mix of people and how they incentiv­ise them (out­side of offer­ing loads of cash). And I’d want to know about how these com­pan­ies move bey­ond what their cus­tom­ers say they want now to cre­at­ing break­through products they will want in future. The report doesn’t answer any of these.

Shame.

When Chev­ro­let jumped into con­sumer gen­er­ated media with its ‘build your own Tahoe ad’ there was a col­lect­ive gasp of ‘you don’t wanna do that’ from the inter­net. And yes, pretty soon people put their satir­ical skills to work with a series of ads high­light­ing the Tahoe’s woe­ful green cre­den­tials and the gen­eral per­ceived stu­pid­ity of any­one con­sid­er­ing buy­ing one. As of right now, you can see over 70 of these on You­Tube (click here for a look). And, as is the norm, once the ini­tial furore died down, it all went quiet.

Well, Wired has fol­lowed up on the story to see how well (or not) the cam­paign actu­ally worked.

As it turns out, not badly at all.

BY ANY OBJECTIVE MEASURE, the Tahoe Appren­tice cam­paign has to be judged a suc­cess. The micros­ite attrac­ted 629,000 vis­it­ors by the time the con­test win­ner, Michael Thrams from nearby Ann Arbor, was announced at the end of April. On aver­age, those vis­it­ors spent more than nine minutes on the site, and nearly two-thirds of them went on to visit Chevy.com; for three weeks run­ning, Chevyapprentice.com funneled more people to the Chevy site than either Google or Yahoo did. Once there, many reques­ted info or left a cookie trail to deal­ers’ sites.

I guess the kind of people who are con­sid­er­ing buy­ing a Tahoe really won’t be that wor­ried about its eco-credentials. They are not mak­ing a decision between a Tahoe and a Prius after all but prob­ably between Chevy’s option and an even worse gas guzz­ler from another man­u­fac­turer. In that con­text, the dif­fer­ence between a brand that helps them cel­eb­rate their decision and oth­ers who remain detached is pretty clear. It’s just a shame that it’s work­ing to put yet more CO2 into the air.

Whether it’s tip­ping points, chasms or tor­nadoes, there’s always been a keen desire among mar­keters to be able to pre­dict the tra­ject­ory of a new product. But, of course, doing so is tricky (to say the least). While describ­ing what’s already happened is inter­est­ing (as the Tip­ping Point showed) apply­ing that ana­lysis to future scen­arios is the real deal.

Well, in a new research paper, two aca­dem­ics from Whar­ton School, Uni­ver­sity of Pennsylvania, have developed a math­em­at­ical model that pre­dicts the pace of new product adop­tion. Warn­ing: the paper is not for the faint-hearted, it is writ­ten with the dens­ity of a small star. But the inform­a­tion within is pretty interesting.

Many of us in tech­no­logy mar­ket­ing under­stand the role that the ‘know­ledge­able friend’ plays in a tech­no­logy sale. These are the early adop­ters, those that oth­ers turn to for advice. These days they are also the ones who blog about their exper­i­ences, con­trib­ute to price-comparison sites and get involved in user for­ums. As such, they are an incred­ibly import­ant audi­ence for new product launches. This paper terms these people influ­en­tials and those that fol­low their advice imit­at­ors.

Chris­tophe Van den Bulle, one of the authors, comments,

“With our model, man­agers can see what will hap­pen to sales depend­ing on the size and beha­viour of each seg­ment. Also, man­agers do not need to guess these unknown para­met­ers; mar­ket­ing ana­lysts can estim­ate them from data about older but sim­ilar products using stand­ard stat­ist­ical software.”

While this sounds a little optim­istic to me read­ing the paper, the the­ory is attractive.

Inter­est­ingly, the paper shows that the chasm the­ory is not the only option. The authors con­sider scen­arios where adop­tion can dip between the early and later phases before sur­ging for­ward. Also, adop­tion by influ­en­tials will not neces­sar­ily dimin­ish rap­idly after launch, so mar­keters may be miss­ing a trick by swap­ping atten­tion over to the main­stream imit­at­ors. And determ­in­ing who is an influ­en­tial and who is an imit­ator is not neces­sar­ily lin­ear (ie it’s not down to who adopts early or late).

The authors go on to look at the two-step flow pro­cess in more detail as well as the tech­no­logy adop­tion chasm. They also talk about scen­arios where imit­at­ors only imit­ate influ­en­tials, where imit­at­ors only imit­ate other imit­at­ors and where the two mix. They then lose me with a series of impress­ive look­ing, com­pletely incom­pre­hens­ible equations.

As I say, come at this one with a large cof­fee and a dic­tion­ary but there’s some inter­est­ing food for thought in here – now if any­one knows a good stat­ist­i­cian who can explain the rest to me…