Irreverent Thinking

Back after the summer (what summer!) break, it’s a great time to be around - the world’s financial markets are crashing, bad advice is coming home to roost and solid thinking is back in fashion. Ace.

I registered my I-name (=dower) today to simplify my OpenId world and it got me thinking back to the future of the web. We are living more and more in an information age but we are limited by our human ability to sift and transfer information and this seems crazy. My PC and the various places (pods) that I inhabit on the net has lots of information about me, letting potential suppliers (retailers, insurance companies) see this information makes a huge amount of sense; they could better tailor their offers and even (on a mass scale) create products and services just for people like me. So why don’t they?

TRUST. No-one seems to trust anyone. So lets go create a framework for a trust network? It seems to work well enough in the social networking world so why not across all web-platforms? OpenID and I-name seem to be part of the solution, a form of machines trusting machines to share the right data at the right time, I’m sure it will evolve once the in-fighting within the various technical steering committees is resolved.

Ignoring (for the moment) the nay-sayers fears about abuse, privacy and control it would certainly make my life a good deal simpler if my (common and public) information was available to every web-resource I visit so it could tailor my experience. Better still, if the i-brokers could provide real secure trust in the same way that credit-card companies do online then you as a i-surfer could choose your provider based on personal preferences. In this world of mass-collaboration it seems a shame that machines cannot yet collaborate in the same way us humans are starting to.

A world where every web-resource knows exactly what it needs to know from you to give you exactly what you want. Sounds like nirvana to me

This week sees the launch of Cuil amid a blaze of statistics (3 times the index size of Google) and a strong side-swipe at Google’s complex popularity index. I’m going to give it a try for a while and the initial thoughts are good. It works well and despite heavy loading on the servers initially it performs pretty well and certainly looks a lot more up to date, making Google look a little outdated in terms of feel.

The usual issue of a lack of a region (UK) specific search is a little annoying but if it takes off (and one VC has just dumped $30m into it) then we’ll see UK versions in due course. It didn’t start well for them, on Sunday when the service went live it fell over almost immediately but that was, according to internal sources, due to the unexpected load on the server. Surely that is a great thing and not a failure? Buying hardware to provide more computing power is cheap.

So far so good, I like it

Tags: cuil, google, searchengine

As we struggle with our new email implementation I’ve been looking at the whole area of how we communicate, both inbound and outbound. Whilst hunting around my attention turned to Flock, a browser I have used for the social-side of life. It offers a neat RSS and social portal (Facebook etc) interface and with some careful configuration I’m testing using Flock for the vast majority of my inbound and outbound communication. In fact, this post is originated inside Flock and nowhere near our WordPress server!

I’ve ranted on before about trial & error economics before and it’s good to see the wider-world start to realise the power in testing. Noone doubts this but the unique flexible, adaptable and changeable nature of the ‘Net means doing lots of tests really does make a difference. Recently we embarked on a rigorous T&E program, 101 tests in 101 days and the outcome was, unsurprisingly, predictable. 90% of the changes generated a 10% improvement and the big 10 ideas generated the vast bulk of improvement - in fact just 2 changes generated over 80% of the improvement in conversion rates.

Over the 101 days we saw a 52% increase in conversion from visitor to data on the site in question. Surprisingly (for us and the agencies involved) nobody predicted which of the changes would be “killers” which reinforces the concept that you have to test and test and test and test even if the results are to establish what doesn’t work.

Testing on the web is simple to do but many people make basic errors in statistical understanding or linking unrelated events and successes. It can be quite dangerous to make any assumptions about the correlation between the tests and the results unless you have a complete understanding of the data and the factors that could and do affect the test.

Bryan Eisenberg is running a series of webinars that, if knowing Bryan and Future Inc, will be pretty interesting. I’ll be watching closely.

MIT has published a great paper, if a little technical, on using visitor focussed content and how to apply customised content to differing personality types. The authors had tried out this approach with a test on one section of the BT Broadband site and had come up with a “20% increased propensity to purchase” and valued this at $80m for this client alone.

Nice reading and kind of validates that we are doing with VITES.

I saw a great post on Digital Design Blog from Tim Barnes at Razorfish | Avenue A. It really is worth a read and about half way through it I had a real double-take moment…..Tim is describing exactly what VITES/Passport Server does. It’s a bit of a shock and a cool feeling to see huge wheels in the big players identify a big opportunity that we have plugged! Just a shame that he didn’t make mention of little ol’ Connected. Anyway, thanks Tim, I owe you a beer.

Unashamedly lifted from Jeff Lanctot (original is here) as it is so of the minute, so topical, so right on. I’ll make myself feel better by saying I am helping to spread Jeff’s wise words

—–

Agencies: Change or Die

May 6th, 2008 by Jeff Lanctot
I’ve recently listened to and participated in several conversations regarding the evolving role of the ad agency. Most marketers and agency types seem to agree that a radical change is necessary. I agree, but I think that change is going to be a harder one to make than many realize.

For many agencies, incremental change feels comfortable. It tacitly acknowledges the greatness of days gone by. It politely suggests that they can get even better. It gently nudges the business forward. The leisurely and familiar pace with which agencies change is also the very thing that threatens their future.

The most effective agencies must live squarely at the intersection of marketing and technology. They should be sprinting to that intersection. Not just digital agencies, but all agencies. As all media becomes digital, any agency that doesn’t view itself as a technology company should commence reflecting fondly on the good ol’ days. The road ahead is likely to be significantly less fulfilling.

Agencies need to make this shift in order to not just lead, but to stay relevant at all. The most engaging customer experiences will be created by designers working collaboratively with developers. The most powerful campaigns will be delivered by creatives that embrace social networks and media planners that understand auction theory. And the most successful agencies will rely on data platforms that give marketers a single view of the customer across all marketing touchpoints. Increasingly, technology is inexorably linked to marketing.

Of course, consumers have already rushed ahead, embracing new technologies without the apprehension and hand-wringing that has marked agency dalliances with the digital world. From Tweens to Seniors, consumers are sharing, publishing and collaborating. They are the new voice of authority in the marketing world- a voice that once exclusively bellowed from the windows of Madison Avenue. Just as they must embrace the role of technology in the business of marketing, agencies must now view customers as colleagues. The passive “target audience” of yesterday is now an active participant in shaping how brands are perceived.

There’s no middle ground for today’s agency. Make small, incremental changes and comfortably drift toward a date with irrelevance. Or make the hard choice to radically reinvent the industry, with the promise of becoming more valuable to marketers than ever before.

—–

End of post

As most of us know, the cost of advertising online is generally going up. We also know that different modes of advertising generate different conversion rates. We also know that demographics can form a significant element in conversion rates. It’s all rather depressing really; we all looked for transparency and a complete understanding of our advertising streams in this new Internet world but, in reality, it is just not happening in a useful way.

Problem Case Study

We have a client who pays a maximum “cap” on his advertising contract with a major advertising network. The overall spend is pretty substantial and runs to a few million per annum. He is facing increasing pressure on advertising budgets as his “cost per acquisition” is starting to spiral. With the advertising network placing millions of adverts in a blind network, you have to think that there are ways to improve the overall performance; one such way is matching the demographics of the sites used with the ideal demographics of the client. This is a good idea, except that demographics are not that accurate and, in our example case, the client gets over 20% of his business from the “unfavourable” 1/3 of the demographics. By cutting out these “undesirables” he would potentially see a 20% drop in turnover. But then again, these undesirables cost over 3 times as much to convert to customers than the desirables. So, how do you solve this problem?

Lets just break open this can of worms and see what the challenges are:

1. The client uses a blind network to advertise on, and can put a “stop” on poorly performing sites. This would drive down the overall cost of the advertising, the average cost of sale (whatever that means in this context) AND the total sales. So, 2 good things and 1 very bad thing.

2. The client has the problem that 20% of his sales come from undesirables and these cost 3 times as much to convert as desirables. Is this because 2/3rd of this sector are wasters or is it simply a case of requiring a far higher level of investment to get them to become customers?

3. The increasing cost of acquiring undesirables is making it almost impossible to increase the advertising budget to the desirables.

4. Their website converts at a staggeringly good rate (over 10%). Is this high conversion rate needlessly driving up the cost of conversion by having to pay for more undesirables?

5. The client has a Cost Per Acquisition contract with the agency and this contract gives a (broadly) fixed rate of acquisition, regardless of subsequent conversion into sales.

Hopefully you can see the dilemma here. But there are 2 alternative solutions to this:

1. Move the agency contract from CPA to CPS (cost per sale) - this drives the agency to generate not acquisitions but actual sales. This is, classically, a good solution as the agency is then driving at the same end-goal as the client but it has its own set of problems. Shifting the risk of converting acquisitions from the agency to the client means the agency will require a greater return on the investment. It also means that the agency has 100% visibility of the success of the client and therefore needs to be an integral partner and not a supplier; not every agency or client is comfortable with that level of collaboration and it can be potentially open to some “over selling”. It’s a difficult track to walk down but for most clients this is the only option open to them.

2. Provide a different experience for differing traffic streams. In other words make acquisition a far simpler task for visitors that match the demographics and a far tougher task for those that don’t. This gives you the best of both worlds. If a “perfect” visitor arrives at the site it becomes incredibly simple for them to be pushed through the process because a) they are more likely to respond to the sales messages do so and b) they have a far better chance of converting to a sale and so offer greater value to the client. If an “imperfect” visitor arrives at the site the acquisition process is more long-winded so it weeds out the “less than motivated” and also ensures that when the acquisition process is complete the client knows a great deal more about them so can better target the sales messages used in the future.

Makes sense? I thought so. We are going to try this with two clients who spend in excess of £5m on online advertising.

As an aside (sort of), Acorn provide an embedded web-service that could pull one of 50 or so demographic types that exist (according to them) in UK. All you would need is the postcode so for site visitors you know little about can suddenly be seen in a wonderful techicolour clarity on the acquisition of a postcode. We’ll be trying this out as well.

As organisations become ever more fleet of foot it becomes clear that the “old world” approach of making huge, long range policy decisions is no longer appropriate. For some time now I have been advocating a “test, test and test again” approach mixed in with a quasi-Kaizen methodology when it comes to improving the performance of websites. This is borne out of being able to have a complete view of what happens when we make changes to a site (thanks to our VITES™ platform); when combined with statistically correct sample sizes and A/B testing, this gives you the clearest (probable) indicator of how successful a change is.

I note that a search on Google turns up some interesting results, with some companies, such as Nike, embracing the idea, and no less than Forbes Magazine citing it as the reason for the wealth of creativity that comes out of our US cousins across the pond. Thomas Edison was once quoted as saying “I HAVE not failed. I have just found 10,000 ways that won’t work.” This approach has its problems which need to be addressed, but the greatest problem is with its name- Trial and ***ERROR*** - which states clearly that mistakes will be made. Sadly most of today’s inwardly-focussed corporates seem to lack the courage to take this step. Even worse are the companies that make sweeping changes without trialling them first; in the case of AOL a $75m loss in revenue.

The companies that do test fly; the ones that don’t end up stagnating and die.

The moral?

To succeed you have to fail occasionally

I do love the way the web is moving - I’m now a full-time convert to Mahalo, why not add me if you sign up to the social bit of it (you can find me here). It does nicely show how the world is moving on and our measurement of trust is changing. I wrote back in September about trusting your Pod, or group of “knowns” (you can read my original post here) and with the emergence of social bookmarking we can see that take another step. I recently joined Diigo, which offers a way to not only bookmark pages but also annotate those pages both for yourself and for anyone who trusts you.

You can see my profile easily and see what I have saved and tagged, or at least the public version of what I have tagged. Diigo call it social annotation which is an interesting concept as it opens the world, and specifically the corporate world, to semantic trust. I keep harping on about e-PR but I wonder if KFC know that I have commented on their “dietary guide” in a less than positive way? The reality of this open (trust) network is that the large corporate animals either better be on their best behaviour or better keep their eyes wide open.

(Thanks to Dan Otterburn for putting me onto Diigo)