E-visionaries: Andrew Hart is the man behind search engine Ask
Jeeves. He spoke to Paul Mason about the challenge of making the
Internet truly interactive - and the opportunities for dotcoms in
today's harsher trading environment
If the Internet is the new rock and roll, then the Home House
club on London's Portman Square will probably one day be revered
like Liverpool's Cavern. As I arrive with Andrew Hart, ceo of Ask
Jeeves UK, some of the best known faces in the UK dotcom sector are
there doing business over breakfast. And - from the high fives and
discreet nods he gets - it's clear he's one of the in-crowd.
But Hart's background is unusual for a dotcom chief executive.
At 34 he already has one big management achievement behind him in
the bricks and mortar world. As managing director of Sunday
Business he was credited with turning the newspaper around after
its chaotic start. Before that he was senior vice president at the
European operation of publisher IDG.
Best known for its "natural language" search engine site, Ask
Jeeves has now moved into providing software and services for other
sites - for example Barclays' B2B.com.
Personalisation is a buzzword for many e-businesses - from
consumer portals to e-procurement exchanges - as they struggle to
differentiate themselves in a crowded market.
But Hart thinks personalisation is old hat. "It's the last
wave," he says. "Why do all these people worry about
personalisation? It actually comes down to the fact that we are
moving from a period of mass media to one-to-one media. That strive
to service the individual is very positive and very good - it's a
very human thing to do. However, it's one very small component in
the process.
"Everyone said they want to do this but they've been limited by
technological processes. Technological processes are very
structured and are focused on the way the computer works rather
than focused on the way the people work. People are very complex -
businesses are getting confused between personalisation and
prediction.
"We predict how someone is going to behave given past behaviour.
For instance I know that you like coffee: next time I meet you I
might assume that you will take another drink of coffee, but you
might not. You might want a beer. Or an orange juice or a water.
But given prediction models, I would know that you like coffee and
I will offer you coffee. If I get it right, you are going to be
very impressed. If I get it wrong, you are going to be fed up -
it's very presumptuous of me.
"So prediction is good when it's right. And what everyone is
talking about at the moment in terms of personalisation is
prediction. They are in the business of learning your behaviour
online and then predicting. They are making it better and better,
because the more you use the Internet the more they can build a
model of it."
However, Hart thinks the industry is nearing the end of this
wave of personalisation. "Prediction is one small component: the
next wave is humanisation.
"Humanisation has many factors: we haven't cracked them, we
don't see what all of them are, but prediction is just one of them.
However, there is one way I can get it right. There is one very
basic human trait that you can leverage, which everyone has
forgotten about: the basic human adventure through dialogue.
"Language is very ambiguous - people aren't very good at using
language. To take a good old newspaper analogy: why have we got so
many sub editors? If people were great with language, you wouldn't
need them. And that's for people who are trained to write.
"So 'disambiguation' is one of the keys to successful
humanisation. Dialogue reveals individuality. As soon as I
understand about you, I can then help you and deliver you stuff
that you want.
"I would say the difference between personalisation and
humanisation is that personalisation is knowing; humanisation is
understanding."
All this would be great if computers were like Hal from 2001: A
Space Odyssey - with the intelligence and bandwidth to behave like
humans. Unfortunately we are years away from that - so just how
"humanised" can the flat screen, narrowband Internet get?
"First of all, the Internet is not a PC. The Internet is a
protocol, which enables many devices to be connected and enables
humans to interact with these devices. It doesn't matter which
device: the issue for the individual is for the point of access to
be a humanistic process. And I think what we're talking about here
is processes. We're not talking about cyborgs or androids or trying
to create them online. Or if maybe there will be some sort of
polygraphic person as in Star Trek who will stand there and look
like a human: that's not what we're talking about.
"Today, code is linear, and humans at the moment are forced to
adapt to the technological processes: think about the VCRs and how
they're programmed. But each adaptation is an inhibitor: an
inhibitor to people relaxing, to people feeling secure rather than
insecure, people feeling at ease with their environment.
"Another problem about humans that technologists forget, is that
most people don't know what they want: they find it very difficult
to express themselves. They don't really know what they want to
buy. So if you're putting them in an environment where they're not
relaxed, where they're confused, it's not really going to
work."
Hart adds that one of his pet hates on the Internet is avatars -
animated human characters that are supposed to promote interaction,
like Miss Boo on boo.com or Ananova on the Press Association Web
site.
"Avatars actually dehumanise: they reinforce the idea that you
are interacting with a computer", he says. "They are trying to put
a human face on technology, not humanise technology itself."
Give B2C more time
Why has so much B2C e-commerce bombed? Isn't the current
narrowband environ- ment, as JP Morgan analysts recently put it,
the "worst possible environment" for B2C?
On the contrary. Hart thinks it has been "fantastic". He cites
offline catalogue marketing: "That's two dimensional - it's a book!
That's even narrower than narrowband. Yet GUS is one of the most
successful companies: they still do it and make a fortune. How can
anyone turn around and say narrow band is the worst possible
place?
"Let's be realistic. How old is e-commerce - forgetting about
EDI - one year, two years? How dare these people go around and make
a judgement on e-commerce that is already bringing in $2bn of
revenue and say it's a disaster or a disappointment. What we've got
to do is get realistic in our expectations of how to develop this
market.
"Supermarkets are fantastic. They're a great place to do
business. They took 50 years to catch on and all they were
replicating is something that is ancient: which is a market, a
street market.
"It is analogous to the move from narrow band to broadband.
Portability went up when more people had cars. Time efficiency was
a problem because people thought: I don't have time to do all the
stuff I like. And there was a cultural shift with all these other
factors going on, away from the need to have that very personal
service you used to get when you went into the green-grocer. What
came instead? A kind of specialised grocer's within the supermarket
and it took 50 or 60 years to replicate an old fashioned
process.
"So why do we now expect that the Internet should be there in
its entirety yet? Let's be a little more patient - it's not going
to take 50 years."
Security and trust
Hart cites industry research showing the big three inhibitors to
online trade: security, ease of use and brand trust. He believes
humanisation can make progress on all three.
"As soon as users say security, we in the industry start
thinking credit card fraud. There are other ways of thinking about
this. Another side to security is the factor: 'I don't feel
comfortable in the environment I'm in; I haven't adapted to this
process because it's a technological process. I'm not sure what I'm
doing; I haven't got used to it.' What happens when you go into a
store that does that? Even if it's a brand new store, you
understand how stores work. You still wander around. You get the
bill. You become acquainted. You become relaxed, then you go out.
The same thing is going to happen on the Web."
Hart recently bought a digital camera: after hours of research
on the Web he was more confused than when he started and ended up
going to a shop where a salesperson talked him through the options
and closed the sale.
"Because Web sites have been designed by webmasters,
technologists and even marketeers, - not salespeople - they miss
that whole sales process out. And so the first step into
humanisation is automating sales. When I'm buying, I want to feel
that you understand me, that you care about me and that this
trusted relationship is going to continue."
So will all this change as broadband access and pervasive
computing arrive?
"The way you posed the question is very much the way the
industry does. Technology outwards. Let's start off with people and
work inwards: they won't actually care what the enabling technology
is. What they will care about is that, wherever they access this
technology, it fulfils the process in the most friendly way.
With broadband, says Hart, the travel industry will be able to
offer interactive tours over DTV. "But if you're going to rebook
the same holiday as last year you don't need the broadband
experience. You can do it narrowband."
What about mobile - what are the possibilities once it starts to
work properly?
"Mobile does work properly," replies Hart bluntly. The debate
over Wap usability is irrelevant, he believes. "Mobile works
properly to do what it's meant to do, and I'm talking about
existing mobiles. Put your phone in your hand; talk into it and
listen to it: it does that very well. We can automate and
personalise verbal conversations now - the problem is not software
but the cost of microphones needed for voice recognition."
Investment tight
We move on to the commercial environment. Ask Jeeves UK has been
set up as a joint venture between established companies - a
commercial model that is gaining ground on the traditional venture
capital route to dotcom glory.
Does Hart think there is too much dumb money chasing too many
dumb ideas? "I see far more portfolio management from venture
capitalists than smart investment. They're happy now to get returns
from one in 10, and lose money on nine."
Many in the dotcom community believe the smart VC money has not
really arrived from the USA yet. But Hart says: "There's a huge
amount of money in this country at the moment. The problem is
getting people to part with some of it - especially venture
capitalists. The people who work for these investment companies are
seeking security just now and this has nothing to do with business
fundamentals: this has to do with the erratic nature of the
market.
He describes a conversation with a well known clicks and mortar
chief: "His stock prices have absolutely tanked: he has no idea why
and neither do any of his bankers. This is scary. The whole market
is out of kilter"
Hart believes a whole series of market conditions have coincided
to produce uncertainty among investors. "We don't know if we're
going into the euro or not. The oil crisis is always emotionally
unsettling in Western culture - it makes people behave strangely.
And you've got all these things going on: mad things happen.
"We saw what happened in Russia, Asia and Japan. People are
scared that's going to happen here. People actually will these
things to happen. And in that environment, VCs are going to be less
likely to hand over cash however good the idea is. It's a little
harder than it was six months ago when it was relatively easy.
"
He believes that within bricks and mortar firms there is a
cultural problem that prevents them seizing the opportunities of
e-business.
"Traditional companies have fairly set processes. Why has the
publishing model hardly changed for the past 30 years? Because the
process is set. The workplace is pretty much optimised. You can't
improve it. What we see here is that, in these companies, people
are brought up in it, caught up in it, absorb it and that becomes
the culture: that's the way it works.
"Whereas in our environment, a young environment, there are none
of these systemised processes. And because the technology is
adapting, we can create them and change them at speed. At Ask
Jeeves UK there were two of us: now there's 87.
"We had a major problem, which was a wonderful opportunity,
which was so many ideas, so many things we want to do and such
limited resources. We had to start putting systems and processes in
place. And we have two people - we call them programme directors.
All they do is sit there, analyse and prioritise what we are and
what we aren't going to do."
Isn't that what we used to call management, I ask? "They are
management but they are not seen as that by people in the company.
They are people with great human skill so that they can work with
every department. They clarify, documenting, regularly analysing,
right across all of them. And they bring order. And all of a
sudden, we've become systematised - system oriented."
Hart believes the biggest challenge for traditional small
businesses is to understand and unleash the power of the Internet
to improve productivity. He says we are moving towards "the end of
non-inflationary times" and that, if small and medium-sized
enterprises don't start getting productivity gains out of the
Internet it will increase the threat of inflation.
"But SMEs need significant gains in productivity," he warns.
"Big companies understand productivity, and work hard on measuring
it, through balanced scorecards, for example." But SMEs will not
make the investment unless you can show them big gains in
productivity, Hart believes. For him this comes down to "stuff you
can automate and people you can focus".
He cites the example of the pub trade where average staff
turnover is six weeks. "If you can get an extra three weeks
retention, and shorten the time taken to train people, that is a
boost to productivity," he says.
A 20th Century thing
Ask Jeeves UK is unusual for a dotcom in that it is targeting a
specific national market. Does Hart think the push to personalise
and humanise the Internet will make it fragment towards local and
national cultures?
"I'm not going to answer yes. Why do we need to fall into the
trap of segmenting people into homogeneous groups, as we've always
done? Why not treat them as individuals? If I went on a UK Web site
and asked the question: 'Are there delays on the tube?' we can make
some great assumptions at first: UK Web site, tube, delay - [that
means] London Underground. From a US access point it could it mean
something different.
"Customers segment themselves, rather than us segmenting them.
It's a very late 20th Century thing: treat me as me, don't treat me
as an ABC 1 married male living in central London. Once you treat
them as individuals, it's the ultimate localisation.
"I behave differently in one location from the way I do in
another. There's a fundamental difference between a Londoner and
someone from Buenos Aires - fundamental cultural differences. But
considering the same two people's family situation, as fathers, I'd
bet we'd become remarkably similar: in fact more similar than
myself and a single man from London."
One part of the UK Government's strategy is to use the Internet
to overcome social exclusion. Does Hart think segmenting and
targeting customers will militate against this, allowing corporates
to discard the "bottom" end of the consumer market? While many
business people might start casting worried glances at their PRs at
this question, Hart strides confidently into the political
minefield.
"The Internet is going to solve the crime problem and it's going
to solve the drug problem. There tend to be pockets of crime,
primarily petty crime and drugs, in less privileged communities.
And if you actually start analysing those communities, there's a
fundamental break-up of the traditional family unit.
"What I mean by that is not necessarily the nuclear family, it
doesn't matter if they have two parents, or whatever. I'm talking
about the sharing, the emotional stability, the communication
between parents and children: the generation gap. And as time moves
forward, the generation gap spreads.
"We all used to joke with our parents when we were kids, about
how different they were. Our parents were probably the first
generation that were aware of a generation gap. And it's going to
get wider and wider with our kids. Technology is one of the causes
and mass media is one of the causes.
"So there's a bigger generation gap. In these communities, the
parent or parents tend to be poorly educated. That can mean they're
poor communicators, not in touch with their emotions. And once
you're a poor communicator and badly in touch with your emotions
and in deprived circumstance: frustration, anxiety and
psychological aggression comes through.
"Obviously that impacts upon the children. And we know from all
the nature and nurture studies, that it's the environment that
determines outcome.
"So what happens when a kid tries to learn - which is a very
natural human instinct - 'Dad, mum, why are rubies red?' Answer:
'Don't be cheeky.'
"Now, in front of the television the only conversation is: 'pass
me the remote control'. And that's not just in more deprived areas.
It's a cultural problem: it inhibits communication skills.
"How do we solve this? If we can help protect defensive parents
from their defensiveness, we can give them with something that
enables them to actually partake to help and encourage the
kids.
"The Internet can give parents a platform to protect themselves
against their lack of knowledge and against poor communication, and
provide a platform where they can interact.
"You can do that in front of a PC, or even an interactive TV
environment. So it becomes: 'Dad, why are rubies red?' 'Let's go on
the Internet and find out'. You can type it in - you can even get
the spelling wrong because we've got a spell check engine.
Making assumptions
"With [Ask Jeeves'] natural language, we make big assumptions
about what people are trying to ask. We never set out to do it: we
just realised what we had afterwards. So in this environment you
are reinforcing positive education, reinforcing communication and
are having positive shared experience.
"It's going to take a couple of generations: we are not going to
solve it overnight. But if you reinforce positive learning
experiences, the kids then enjoy learning. They go to school, the
same type of thing happens - they're working with technology, it's
cool, it's there, it's orderly. And all of a sudden you're back on
the path to learning. And now you're not the geek who goes there:
everyone's there.
Next week we profile Michael Edelman, European head
intellectual property exchange yet2.com
CV: Andrew Hart
CEO, Ask Jeeves UK
Formerly managing director of Sunday Business and senior vice
president in EMEA for IT publisher and research group IDG. He has a
degree in Economics and Accountancy from Manchester University.
Aged 34, he joined Ask Jeeves in January this year.