Personalisation will fashion the boutique future for online travel

The emergence of the ‘personal web’ promises to be a boon for travel professionals. What can travel learn from the world of fashion? Lee Hayhurst reports from a QuBit round-table

The emergence of the ‘personal web’ promises to be a boon for travel professionals. What can travel learn from the world of fashion? Lee Hayhurst reports from a Qubit round-table

Forget what you have might have thought about the future of the web being all about robots and algorithms. More human, personalised recommendations are on the way and serving up those will require knowledgeable travel professionals.

It’s easy to imagine a world where computers know you and decide what you might like. But a Terminator-style ‘Skynet’ of artificial intelligence still remains fanciful science fiction.

So what’s in store? London-based web analytics firm QuBit believes that as the techniques of personalisation advance, the need for human input into recommendation engines will become more vital. The reason, says managing director Graham Cooke, is that firms will need experts to make sense of the masses of data they have to ensure different customer segments are given the optimum experience.

That means using experienced travel professionals to fine-tune the machine algorithms, improving the relevancy of the offering and therefore boosting conversion rates.

This would exploit insights such as what products go with what, complementary destinations or activities, and the behaviour and needs of particular customer segments.

These can be pumped into the online sales and marketing process to refine campaigns and avoid the more blunt ‘law of averages’ approach of purely data-based arithmetical calculations of likely behaviour.

“The change on the web today is that you have utility type businesses and experiential,” said Cooke.

“Utility is about being able to find that product as quickly as possible – a very clinical experience. Amazon is going to represent 25% of all transactions online by 2016 and are a firm owner of that position.

“On the experiential side you have the emergence of many brands selling unique product and have a unique proposition. They control their supply chain and control that experience for the consumer.

“We see this sector as one that’s really going to flourish and emerge.”

One example given in travel was the boutique hotel specialist Mr & Mrs Smith, with its curated collection of hotels and loyal following.

But examples given from the world of fashion included Fab.com, which drives 50% of its traffic from social media, and Net-a-porter, which has moved into publishing in a major way to use content to drive conversion.

Cooke said: “We will see a lot more boutiques and brand-driven boutiques out there. It’s about finding problems [the consumer has], delivering solutions and delivering quickly.”

Cooke said sectors such as travel, with tens of thousands of product options, had to find ways to help people navigate through the choice.

“This becomes a data-driven problem and you’ll need to mix that with a creative solution if you want to provide different people different ways of navigating through that information.

“Some want an interactive recommendation, others may want a social recommendation. Businesses such as Airbnb and TripAdvisor are integrating that social layer straight into the website.

“What drives recommendation? If you have thousands of products you need to help people through that.

“Product recommendation algorithms do that to an extent but it’s difficult to make sure those algorithms get it right. In some ways human experts trump algorithms.”

QuBit expects that future generations of Content Management Systems, developed over the next five to 10 years, will invent new ways for customers to interact with inventory compared to the current ‘online catalogue’ model.

“The world is actually going back to being more personal,” Cooke said. “It’s the personal web, the user web – all about understanding the customer.

“We will see the development in multi-channel and online, where the potential is for the high street to become an extension of your marketing channel.

“Get them in and drive them to your website as long-term customers. It’s cutting through the Big Data noise and getting back to good merchandising, good marketing.”

A recent QuBit survey found 92% of companies were doing some form of web analytics but only 18% of business was engaged with some form of personalisation strategy.

“In the past the necessity [for personalisation] has been relatively low because you could, relatively cheaply, buy new customers and get people to your site at a relatively low cost.

“It was a free run in the early days but cost-per-acquisition has got four times more expensive in the past five years, driven by lower conversion rates – up to 50% lower – and higher cost per lead.

If the market still focused on how to get traffic. The reason we started QuBit in the UK was that the challenge would be greater here.”

“The UK is ahead because there is much greater spend on digital media and much higher spend per capita on commerce. That’s created more of a crunch than in the US. Over there a great chunk of the market is still focused on how to get traffic. The reason we started Qubit in the UK was that the challenge would be greater here.”

Seven top tips on personalisation

1. Employ experts both in terms of the product and the market you are selling in to ensure you have an authentic, expert voice
2. Use your in-house expertise to intervene to prevent customers making errors of judgement, such as booking flights with overly long transfer times
3. Analyse your customer base to see if there are distinct groups that give you statistically significant insights into their shopping behaviour
4. Track your customers as they move through these distinct groups. A new customer will initially behave as a first-time shopper, but will quickly move category
5. Curate your product range so that it fits the needs and requirements of your customer base and helps them hone their choice
6. Constantly analyse what is a barrier to conversion on your site and act rectify it quickly. If this does not work, try something else
7. Create content that inspires and present it on your site to optimise engagement and, ultimately, conversion

Fashion case study: Farfetch.com

High-end boutique online fashion store Farfetch.com brings together 1,000 product lines from a selection of small retailers around the globe.

The firm says it is good at breaking down geographical borders to commerce by the way it sources its product from 18 countries and sells it in 140-plus markets.

Due to the nature of the product its range is limited in terms of volume – and so it is inherently personal to the customer.

Andrew Robb, chief operating officer, said while personalisation was important, it would not go as far as the term from the late 1990s of ‘one-to-one marketing’.

“I think one-to-one marketing is the wrong thing to aim for. It’s very, very difficult figuring out what someone wants without relating them to other people,” he said.

“There is a point at which you drill too far down and it does not work. It’s possible, but it does not makes sense because you do not get the scale out of it.”

Robb said personalisation could give you key insights into large sets of customers, although he warned about leaving this all down to a mathematical equation – “the tyranny of averages”.

But drilling down into conversion rates can reveal some obvious issues that a set of customers is having in a particular market. And these can be easily fixed.

“We are just starting out really trying to personalise the offering,” said Robb. “Our problem is quite acute because we have a broad product line for the sector we are in.

“But at the same time it is quite thin because we are working with relatively small boutiques. Sometimes there may only be four items available.

“As a consumer how do you navigate thousands of products?”

One insight Farfetch has discovered is that men use filtering much more than women, meaning men are more prepared to use the tool to create a personal experience.

Robb said: “Ultimately it’s about pushing product to people who you hope will engage with it, whether they are in search mode or browsing.”

Farfetch has segmented its customers into significantly large enough groups to generate statistically meaningful insight.

It has identified two broad groups that represent 40% of its business, both of which are very high spenders. One is professional women, the other is ‘fashionistas’ men and women who are highly a-tuned to what is trendy.

These groups are treated differently and served with different messages, said Robb.

However, a further complication is that individuals will move around the various customer segments as they interact with the site and modify their behaviour.

Fashion case study: L K Bennett

High street fashion retailer LK Bennett has found personalising its web experience has helped fuel international growth.

Sarah Lukins, head of e?commerce, said the firm has been trialling personalisation with QuBit.

Some simple stuff is being done initially – like making sure shoppers who arrive at the site from international markets are treated differently to domestic ones.

Something as basic as just making sure that international clients are served with an international delivery call-out has been seen to significantly improve conversions.

“We have been looking at this, like how to call out our sizing of products,” said Lukins.

“It’s been beneficial to bring out messaging that we thought was obvious but which was not to certain defined groups.

“We want to get more involved in serving defined content to defined types of people, particularly as we develop a broader customer base. From an international point of view personalisation has been important.”

LK Bennett has launched a US site but found it needed to hire an American to run it because it was important to make the site as authentic as possible for audience in the US.

“There was a lack of local knowledge when recruiting people in the UK to work on an international site,” Lukins said.

“You can tell if you’re being delivered content being written not in the country of origin and by someone who does not know the country well enough.”

The brand draws one-third of its Facebook fans from the US, many of whom had been asking when a US site would be launched, and the market was one of the top drivers of traffic to its website.

Lukins said expansion had been helped by the fact that LK Bennett now has physical stores and concessions in the US.

The brand is now looking at further international expansion in places like France, the Middle East, Asia and Australasia, where it has stores.

QuBit’s Graham Cooke will be among the speakers at this year’s Travolution Summit in London on November 18, where he will talk about personalisation. Nishma Robb, iProspect chief client officer, will also address the event, citing the fashion sector as an example of best practice. Early-bird delegate places are available for just £99. For details, and to register, go to travolutionsummit.com.