I was looking through old customer comments recently stored on our system and one stood out:

“I can’t close because the customer is buying on price even when the cover isn’t right for him.”

This came from an insurance broker in specialist lines business – that means he wasn’t dealing in something really commoditised like private car policies, but a class of business where a smattering of knowledge could go a long way.

It could be that he’s just a bad salesperson – some brokers are still of the opinion that the market still walks in their front door year after year rather than heading to the perfect information of the Internet.

But I think it’s a deeper problem that needs to be addressed now if the general insurance intermediary is to survive.

For those who don’t know, the insurance industry in the UK for “normal” business (i.e. your premium is calculated based on some rules and variables with a smattering of seller adjusting) is conducted directly with the insurer (think Direct Line) or via brokers.  Insurers will give you their rate, often discounted as middleman fees are removed, and brokers will give you rates from a number of insurers, earning their money from a commission.  Since the rise of the Internet and direct sellers, the consumer has been educated to buy on price alone.

The broker has the run of a few insurers and often negotiates special rates for individual niches, so they’ve often been able to work well in this model.  Volumes of business have increased, particularly as aggregators send firehoses of leads in.  Brokers have been given better and better technology with insurers publishing rates on extranets rather than paper rating guides.  Everyone’s been reasonably happy.

But.. few seem to notice that as the demand for computerised rates increases from insurer to broker, two deadly fronts are advancing on the intermediary.

1: Computerisation of all rates closes the gap between intermediary sales and direct sales

If I’m an insurer and I’ve invested a ton of money into computerising my rates, covernotes, documents, MTAs etc – I’m going to eventually start to wonder what it will take to put an online interface to the public, under my own brand or a broker I happen to own.  I lose my distribution chain of an agency network, but I gain a bunch of commission, control and negotiating clout.

Imagine computing power in twenty years time, and the data… I could have an insurance brain making up daily pricing based on current risk, adjusting premiums as data flows in and out from any number of sources.  Claims data, demographics, crime figures, current affairs, sporting events, weather, company announcements, health statistics… the list is endless.. they’re all available in real time streams.  Doesn’t it seem very quaint that your premium is based on data that’s a year old at least and will be fixed for one year hence?  In other words – by the time your policy expires, the data upon which your risk was calculated is 24 months old?  Isn’t that akin to town planning based on archaeological remains?

2: If all brokers are quoting from the same extranet rates for the same products, the only thing left to compete on is <> price.

If you’re going to get the same cover and price from Broker A, B, C & D – what’s the point in talking to B, C & D after you’ve spoken to A?  You probably don’t want to run through 90 questions again and listening to another 10 minute FSA disclaimer is not going to be fun (why don’t we centralise FSA disclaimers?)

The answer is of course to stop competing on price alone.

Coffee shops did it.  Why can’t brokers?  Budget airlines broke in on price, yet are now subtly moving away and adding on new reasons to choose one over the other.  Even utilities companies rail against price only competition now and again with convenience, green features and the like.

Learning to differentiate NOT on price is the only defence against the above.  More later..

I’ve been reading about how people are confused by O2′s pricing strategy for the IPhone 4, so I thought I’d take a look..

These things are designed to be impossible to understand, I’m fairly certain. So I banged the numbers into a spreadsheet (that was hard enough).

Here’s what I discovered…

1: The 18m £209 v 24m £279 for a handset on 100 minutes is a red herring – although the contract costs more over the 24 months than 18, the actual cost per month is £41.61 on 18 months v £36.63 on 24 months.

2: Cost “per minute” is insane on the 100 min contracts, but drops massively at 300m+

3: On a cost per minute basis, there’s not much in it between 18 & 24 months at 600m per month

4: At 600m the 16GB 32GB work out roughly the same per minute cost on 18 or 24 m contracts

5: Mobile phones cost a shit load more than you think.. how did we reach the point where £50/mo is a necessary expense! Wowzer.

Good luck picking your poison!

Here’s something I don’t do very often.. a review.

I’ve been sparked into action by a twitter automation tool, Hummingbird2. Before I begin, I’d just like to point out that I’m not an affiliate and any links are straight, not coded.. so I speak my opinion..

I needed a tool to help find interesting folk on the Big T, so I went shopping and based on what it said on the tin, came back with HB2. Silly me.

The problem is that it bills itself as being a highly accurate finder / adder / unsubscriber based on targeted keywords in the public stream.

It’s true – you find people by telling it what you’re interested in. It finds anyone who happens to have typed or retweeted your word and compares them against a few heuristics to see if they’re up to scratch (simple scoring mechanism based on your settings for min tweets, followers etc).

The problem is that the search engine seems to be dated circa 1995. There’s no support for operators, negatives or even a simple AND. For those who understand such things, you’re basically getting:

SELECT * FROM firehose WHERE msg like ‘%keyword%’

I don’t get it. Even the public API for Twitter supports most of their fancy switches and query language.

So the net result is doing anything clever, like filtering out messages from foreign languages, restricting to a very specific subject, time, place etc is impossible.

I ran it for a few hours and have spent a few days trying to clean up the mess it made of my account. In spite of setting the quality requirement to B+, I ended up following spammers, poker people and *endless* “retweet to win” types.

Another gripe is the unintelligent unfollow system. There’s no VIP list or way of protecting the innocent, there’s no flair to how the unfollowed are chosen. There’s also no way to manually unfollow users in case it picks up someone you’re not interested in. The big song and dance about clever unfollowing is simply “unfollows the oldest person who hasn’t followed you back”.

The makers seem to have spent a lot of time on a nice polished interface, a half baked Twitter client (presumably to give some kind of “free download” and forgotten the most basic things – if I want to employ a software tool to help me follow interesting / targeted people, then it better be able to target.

I did write to the company about my concerns, but I received a very odd email back being all philosophical about success (“Like any tool if you are looking for the tool to make your success you will find that your success is limited.” Er, yes. Now – why doesn’t it filter like it says it does?)

Summary – if you want something to sit there and follow zillions of people in the hope that they follow you back, it’s probably quite good. If you want to find targeted users matching specific criteria – I found it useless.

[tweetmeme]A buddy sent me a link to HEMA’s website today, it’s a Dutch department store. The instructions are “Load page, watch and do nothing.”

It’s brilliant and will be tweeted everywhere. The only problem is it’s quite tricky to navigate through to the store proper once it’s done it’s thing..

Check it out though, I’m sure you’ll RT…

http://producten.hema.nl/

[tweetmeme]The FT has a moderately interesting podcast this morning about “Capturing the Mood” of social networks for brands.  It’s an interview with Kelly Dempsky of Accenture and is an attempt to find out how Accenture are working with their clients to try and “listen in” to social network content and understand the impact of what’s being said.

The angle is more that social content is a PR channel to be managed rather than a promotional channel to sell through as companies find that they’ve moved from “broadcasting a message to becoming part of the content.”

The technology is all very fledgling but the interview lays out what needs to be done:

• Monitor content for positive and negative “sentiment words”
• Put those words in front of humans, machine language processing isn’t clever enough yet at this level
• Measure the reach of the conversation – followers, subscribers, comment levels, history
• Who’s making the comment (who’s saying it – something Al Gore says is different to something Jerry Springer says)
• What network is the comment made in, the flavour of the network for receiving data – positive or negative etc
• The connections that the comment might travel through

All in all it sounds like an awful lot of marketing dollars could quickly be swallowed by the exercises that a large organisation might feel it needs to go through, but done right it could provide a superb finger on the pulse realtime measure of the market’s opinion to a brand.

Marketing VPs and investors take note!

Link to full podcast, c. 10 mins: Capturing the Mood of Social Media

[tweetmeme]This came via email for me today prior to asking me to fork out £195 for a conference seat:

Social Media is great but where is the return on investment?  The point here, is that we are actually asking the wrong question.  Social Media is not a platform, but a tactic which is about relationship building, branding and being where people are looking.

This is exactly the type of message that drives me nuts.  You can’t approach a proper business person and say that it’s wrong to expect a return on investment from an activity.  There must be a return, or you go out of business.

Business exists to make a profit.  Non profits still make profits, they just spend it all on what they’re working on and don’t leave any in the bank.  There must be a return on what you do, otherwise eventually your money will run out and you’ll have nothing left.

A Social Media strategy has a cost, just like anything else.  If you’re an entrepreneur who spends 2 hours a day tweeting, then you’ve spent 2 x Your Hourly Rate… if that’s $50/hour then in one month you spend $2000 on tweeting.  If that makes you $2001 then you’ve made a profit, otherwise you’ve broke even or made a loss.  Simple.

“It’s about building relationships”.

Sure – but why?  How?  I have enough trouble maintaining relationships with my few oldest friends around the world, how can I hope for a pizza brand to create a relationship with me?  It’s not about building relationships, it’s about building mailing lists.  This crap wouldn’t be so hard to swallow if the pushers told it like it is.  You give me permission to try and sell you my stuff now and again and I’ll attempt to amuse you in return with some widget, game or connection.  A modern take on have-a-free-gift-in-exchange-for-receiving-my-mail-order-catalogue marketing.

Far better these endless seminars and agencies would be if they told people how to monetise their social media activity, rather than blatantly advertising, as the above, that they don’t know how.  If I’m going to give money away and expect no return then I write cheques for Great Ormond Street, otherwise, my dollar’s gotta work for me.

End of Rant.

[tweetmeme]I read with interest, and slight glee, that Google is being hit hard by consumer groups complaining that it should have telephone support for the Nexus One instead of the traditional Google way of email only.

As a long time user of Google’s services, I’ve often been frustrated and amused by the fact that the hundreds of thousands I spend with them doesn’t give me access to even having a telephone number there (we used to have one, but it’s been recently replaced with a voicemail asking us to email). In any other company in the world, if you spend half a million or so a year, you’d expect there to be a salesperson ready to jump.

Not with the Big G though, bless ‘em.

Many are also complaining about the amount of time that it takes Google to respond to queries. Google said it would answer problems via e-mail though any response may take a day or two to arrive…. A common sentiment on the support forum was that for the $500 people have paid for the phone they should be able to call a dedicated help line. bbc.co.uk

They’ve brilliantly and efficiently changed our expectations, in part through being what is effectively a monopoly, of customer service. Granted, their email response is pretty good these days, but nonetheless seeing how a new clutch of Google customers is reacting to dealing in the Google Way is a reminder that there’s still a long way to go before every buyer can be beaten down into using the “new” ways of ecommerce.

I wonder how much we leave on the table by excluding them in what we do; or more importantly what the opportunity will be to pick up the ball of old-style customer service in the new-style economy?

I spend a ton of money on PPC for various things and spend an awful lot of time analysing data, and the last few weeks have been a perfect example of how not keeping calm in a market can lead to havoc.

In the run up to Christmas, demand for this particular niche actually drops off and does so quite significantly.  Most of the year things are pretty constant at $N per click, but as demand naturally dropped off in the last few weeks, I saw our placement drop from top three to second page, with bid requirements for returning to our former glory shot through the roof to 2 x $N and above.

I consulted with a colleague in the same industry also dialled into PPC and he said the same thing was going on – advertisers were throwing more money at their PPC in an attempt to buy business that wasn’t really there, forcing prices up, moving encumbants down the ranks and starting a vicious circle.

I’ve chosen to ride it out.  We can live without the demand this month, but can’t live with buying business at a loss (this isn’t the 90′s after all!).

It’ll be interesting to see what happens when the market returns in January and demand literally spikes.  At that point there’ll be loads of advertisers with over the top bids, a slew of clickers shopping around and a lot of runaway budgets.

I know the upturn in bids isn’t a result of automation in most cases too as I know many of the advertisers and how they operate.  It’s a human thing that’s ironically caused by too much, but not enough, data.  If you were to just set and forget your ads, you’d tick along and neither be the best nor worst, you start to do some analysis and think every answer is in the spreadsheet but you miss the crucial piece – common sense.

So – countdown time.. let’s see what the market brings in 7 days time…