Thursday, September 27, 2007

Spend less on advertising!

A version of this piece was published in Marketing in 2007



This week I recommend you spend less on advertising online.

Last year’s IAB report showed a 41% increase in online advertising spend, this year’s figures are due to be released this month - the only discussion is how much the increase will be, not whether there will be one. This is of course very good news for companies that are in the business of buying digital media, so why would I suggest you spend less?  

First, take a look at where you are spending your digital marketing budget.  If you are like most companies, 100% will be spent on display advertising, search and affiliates.  Driving traffic to your web properties is of course very important, but what too many companies are still failing to recognise is that getting the customer to the front door is only one part of the sales process.  Arguably it’s the easiest part too; converting the visitor into a customer is much more challenging.

I’m a numbers guy, so let’s get back to budgets.  If you spend £1,000,000 on online marketing to generate 2,000,000 unique visitors and manage to subsequently convert 3% of your visitors you might be pretty happy. That’s 60,000 new customers at a Cost Per Acquisition (CPA) of £16.66.

But think about the key figure in this equation – the conversion percentage.  What if we could raise that by just a quarter of a percent, by taking £50,000 of budget and using that to focus on site conversion rather than driving traffic to the site? 

The new numbers are revealing.  Using the same ratio as above a spend of £950,000 (that’s £1,000,000 less the £50,000 earmarked for increasing site conversion) would generate 1,900,000 visitors. If you convert at 3.25% you’ll get 61,750 new customers at a CPA of £16.19  So that 3% more customers for a CPA that is 3% lower. 

This is of course a no-brainer.  Who wouldn’t opt to spend £50,000 to increased conversion by a quarter of a percent? 

Well surprisingly, a lot of companies don’t – and it’s not as difficult as you might think.  Many companies say they have this under control - they regularly perform usability testing and adjust their site accordingly.  This is a good start, but it puts the emphasis on the site, and not on the customer.  That’s why Customer Experience testing is an increasingly popular approach as the online channel grows in value.

To understand the difference, think about how a typical usability study works.  Typically 8-12 participants are asked to complete a series of tasks that follow a relatively logical pattern – find information about product X, register, buy the product, contact the company, etc.  The focus is the site, and the key question you’re answering is, “does what I’ve got work”.

In contrast, a typical Customer Experience test begins with a 30 minute discussion before the participant even touches a PC.  This sets the scene, establishing motivations and emotions prior to the start of the purchase process (perhaps a search).  Once on the site itself the moderator will allow customers to take their natural path rather than following a prescribed list of set tasks.

This approach allows the marketer to garner much more valuable information – what the customer actually wants – rather than whether what has been provided works.   

As a marketer, this approach should sound familiar.  Think about how you test consumer reaction to new products.  Your main interest is not whether the product works but rather whether customers want it, how they use it and how they view it.

The difference approaches might sound subtle, but the difference in results is not. If you want to maximise your chances of increasing online conversions, what should be at the heart of your online business strategy – your site or your customers?  A usability study by its very nature puts the site at the heart of the process.  Shouldn’t customers be the heart of your business?

Thursday, September 20, 2007

Google stops agency bribes

A version of this piece was published in Marketing in 2007


Late last week, Google announced the latest changes to their controversial ‘Best Practice Funding’ scheme – the most significant of which is, it’s being dumped.

Launched in 2005 with just 13 weeks’ notice, the scheme upset many agencies and concerned many advertisers, who saw it as anti-competitive, market distorting and untransparent.

The scheme was designed to incentivise agencies to invest in search, by kicking back a percentage of their spend with Google.  But of course because it was only predicated on an agency’s spend on Google, its purpose was to incentivise investment on Google.

But for many agencies who had transparent relationships with their clients, the kickback was passed directly back to those clients, thereby negating the incentive effect and only rewarding agencies who kept the payment.

Advertisers complained that the unpredictability of the level of rebate made it harder to budget, and found that rebates coming back five months after the activity simply got deployed into whatever happened to need funding at the time – as often as not, something other than search.
But much worse than this, the scheme incentivised all sorts of behaviour that the ubermenschen at Google hadn’t anticipated.

There are a lot of clever people in marketing and media, and they’ve spent a large amount of time over the past two years figuring out how to play the system – time that might have more profitably been spent making their search campaigns work better.

Google’s offering an extra 5% ‘growth kicker’ to agencies that showed a given level of quarterly growth led to certain advertisers moving agency every quarter, knowing that their addition to an agency’s billings would qualify it for the extra 5%.  And since this applied across all the agency’s billings on Google, these advertisers often demanded a share of other clients’ rebate too (did you get all yours?).

The scheme’s market distorting effect was proved when tenders for search business started to include the question “what is the level of your agency’s BPF rebate?”, with procurement departments (understandably) seeing this as a part of the competitive dynamic between agencies.  

This seriously disadvantaged small agencies, creating barriers to entry and allowing poorer performing (but multinational) operations to offer greater rebates.  (arguably Google were knowingly subsidising the fees of poorer performing agencies in the knowledge that their inefficiency was actually more profitable to Google).

So the news that from 2009, the scheme will be dropped is a healthy development for the search market, which has grown to over £1bn in the UK.  In the meantime, it’s adjusting the scheme, dropping the problematic growth kicker and reducing the qualifying thresholds for the higher tiers of rebate.  Additionally it’s dropping the flat-rate 10% commission on non-search products like YouTube.

Whilst some agencies will fear the abolition of BPF, relying on it to fund their businesses, the move to a net model will create a level playing field in this market.  Agencies will have to demonstrate the value of their service, and will now only compete with each other on ability rather than rebates.

Advertisers’ attention will be focused on the role and effectiveness of their search activity, and won’t be subject to the distracting lure of deceptively cheap fees and market-distorting incentives. 

And whilst it’s irritating that the scheme will continue for a further 15 months, these businesses do need time to adjust their commercial arrangements, and Google’s preparedness to listen and to learn from previous mistakes is to be welcomed.

Google this year is expected to make more profit in the UK than ITV and Channel 4 combined, and with a market share of over 80% its imposition of the BPF scheme has led to it being accused of abusing a dominant market position.

With the EU’s competition authorities taking industry soundings over their acquisition of DoubleClick, the company is understandably wary of such accusations.  BPF’s demise is to be welcomed, but it could be little more than a sign that Google has bigger fish to fry.

Thursday, September 13, 2007

Brand terms and the navigator

A version of this piece was published in Marketing in 2007


Every day, I walk through a street market to get to my office.  From fruit and veg to improbably large pants, the traders call out their wares, trying to attract attention and bring customers in - but whilst they all look the same, very different motivations drive those punters. 

Some know exactly what they want, and whilst they hear the stall-holder calling out, they’re going to buy anyway.  Others are simply browsing, and only are only attracted to buy when they hear that call.  If the traders had to pay every time they called out, they’d be a lot more careful who they called to.

The last step in the purchase process, search is on the face of it, hyper-accountable.  Tracking lets us to establish precisely which keyword resulted in a sale – allowing the online marketer to pick out which of perhaps thousands of keywords are generating business – and that information can be fed back into bid strategies, continuously adjusting the amount it’s worth paying a search engine for a click.

Attracted by the apparently low cost per acquisition in search, investment has spiralled, often at the expense of other media.

But whilst there is no doubt that search is an incredibly valuable tool, most marketers are working under some fundamental misconceptions about how search works.  And as a consequence, they’re often substantially over-valuing what they get.

Why?

Because almost half the people who search directly for your brand name aren’t really searching at all.

Nielsen research tells us that 43% of people don’t type the web address of the site they’re looking into the address bar of their web browser.  Instead, they type it into a search engine.

So when someone types “easyJet” into a search engine, it isn’t because they don’t know the address is www.easyjet.com.  It’s because they don’t know how to use their browser properly, or because they can’t be bothered. 

Either way, they’re not searching.  They’re using the search engine not to find the website, but as a means of navigating to it.

The impact of this is profound.

Navigators were going to your site anyway.  The search engine added little or no value here – it didn’t present your brand to them when they weren’t expecting it or were considering another – it merely directed them on.

This has some value, but this value isn’t equal to that generated when a consumer is actually searching.  Here, a search engine has moved them on substantially in the purchase process, and often at a critical point.

On Google, easyJet have successfully prevented others from bidding on their brand term and are almost certainly therefore paying the minimum bid of 1p a click.  The term itself is probably their best converting term in terms of sales.

But if 43% of the people typing that term into Google are Navigators rather than Searchers, and we say for argument’s sake that a navigation is worth one third what a search is (I’d argue it’s really much less), then the true cost per acquisition is going to be much higher than the conventional wisdom would measure.

Under these circumstances, the ‘true’ cost per acquisition is actually 40% higher.

That knowledge could make the difference between a term being cost-effective and a dud, and particularly if it’s a brand name that can’t be trademark protected.

So how can we tell a Navigator from a Searcher?  If we could tell them apart, we could choose not to put a paid-for listing in front of a Navigator, letting them rely on our natural search results, and saving a bundle.

But regrettably, you can’t.  They type your brand name and either buy something or not, and you’ll never be able to spot one coming unless you’re a mind-reader.

But you can gauge the overall impact of Navigators, and factor cost per acquisitions to allow for this.  If you don’t, you’re paying the search engine for value it hasn’t created.

Thursday, September 6, 2007

Online travel and the trialogue

A version of this piece was published in Marketing in 2007


Back in 1996, I was standing on a conference platform talking to a room full of travel agents.  I had a laptop in front of me with a dialup connection to the internet, and I decided to take a risk. 

I’d just spent twenty minutes talking to them about how the web might impact on their business, and frankly, they weren’t impressed.  “People”, one delegate said, “will always want the advice only a travel agent could give them”. 

I’d just returned the previous year from a round-the-world tour where every hotel I’d stayed at had been found online, and flushed with confidence, I asked the audience to name anywhere in the world, promising I’d find them a hotel there.

“Easter Island” called out one smartarse at the back.

It took me an admittedly nerve-racking thirty seconds to find one and read out the details.

You could have heard a pin drop.

Now, travel is one of the biggest commercial sectors online.  easyJet sells over 90% of its flights online, around 15% of searches are travel-related, and the European online travel market was worth E38billion last year.

Not bad for ten years’ work.

But although technology gave consumers information about far-away places, access to airline and hotel availability databases and the ability to communicate directly with these organisations (without waiting for the next assistant to be free), it gave them something else - which until recently businesses have not got to grips with.

It gave them access to each other.

Through sites like tripadvisor, holidaywatchdog and myholidayreport, consumers started telling each other about their real experiences, good and bad.

“The manager approached us by the pool, saying the fans we’d bought were using too much electricity… we will never stay at this hotel again”

The second bedroom in the Chateau had a continuous combination smell of mould and rotting flesh."

The sheer granularity of these sites was unachievable before users started creating the content – the daddy of them all, tripadvisor, claims to have over ten million reviews covering 190,000 hotels and 140,000 restaurants. 

Barry Diller’s InterActiveCorp, owners of Expedia, saw the potential back in 2004, when he acquired tripadvisor, placing real reviews next to hotel listings.  In the UK, Thomson asks consumers to submit reviews, but and the company reserves the right to edit, refuse and withdraw contributions. The result is a very different feel to tripadvisor - in 15 minutes I wasn’t able to find a single negative review.

The latest addition to these is a familiar face from the dotcom boom, reincarnated as a travel site.

Boo.com is fast, well-designed, packed with useful features that speed up the experience.  It seamlessly merges data, listings and user-generated reviews and packages it well.  It’s honest with its readers – a one-line review read: “situated as it claims on a quiet street, this hotel is also near a noisy one”.

The old-world view would see this as a high-risk strategy.  Placing bad reviews next to hotels you’re offering for sale can’t be in their commercial interest can it?

But the reality is that consumers are using other consumers’ opinions to rationalise purchasing decisions.  There’s nothing a travel site can do about this, so bringing these reviews into the site achieves the double benefit of increasing trust in the agent’s brand, and not losing that consumer when they go off to check out your recommendations.

Boo crashed and burned the first time round, trying to sell a product people weren’t ready to buy online, using technology that users couldn’t access using the slow connections of the time, and failing to control costs. 

This time, it’s in a booming sector, with technology that puts the consumer at the users at the very heart of creating its product – a true trialogue brand.