With retailers facing the twin challenges of rising consumer expectations and pressure from the likes of Amazon and Alibaba, the need to deliver a truly omnichannel experience is greater than ever. InternetRetailing sat down with Carsten Kraus, AI expert and founder of FACT-Finder, to learn how they can use customer data and smart algorithms to achieve this.
The challenges facing retailers
It is hard to get true omnichannel – most retailers offer multichannel, in which individual routes for customer interaction co-exist but still appear separate and occasionally disconnected to the customer. This is a problem since a retail brand exists in the customer’s mind as a single entity and not as ‘online and offline’ departments. Store operations and online activity have grown separately, and retailers face the challenge of merging working practices, operations and the customer experience into a consistent, seamless whole.
UK retailers often focus on the challenge from Amazon, but European retailers also need to consider how they will co-exist with Alibaba. Alibaba works directly with Chinese manufacturers and can bring their products directly to consumers at lower costs, cutting out the retailer in the middle and shrinking their margins. This is a twin threat to retailers – exemplary logistics and cheap, directly-sourced products.
To counter this, stores must become digitally enabled and integrate with online into a seamless user experience. Amazon and Alibaba do not yet offer physical stores allowing people to touch and look at products so this is an opportunity to differentiate.
The store can be a marketing tool. When you have a physical interaction, the trust is much higher whereas online you only have a two-dimensional view.
How our product can help
Our product, FACT-Finder, helps retailers provide a state of the art online store. It’s not just how it looks, it’s also about the actual experience of buying products. The customer knows what they want and wants to see it on the screen straight away.
Too many retailers are sorting search results on non-personalised metrics such as price. Amazon never does this: they put the products that customers are most likely to buy first. Our algorithms learn from customer interactions, so if customers respond in a certain way to a search result the shop learns this. It also offers semantic learning, which analyses what differentiates the product that got bought from the one that didn’t.
Consumers want to see what’s in stock in their local store so they can pick it up the same day. We do this with big online chains such as electronics giant MediaMarktSaturn and home improvement specialist Kingfisher. We collate the information of which store has what, so that when a customer in a given area searches for a camera, FACT-Finder weighs the local availability with other metrics such as how likely a product is to sell. The customer is then presented with a locally available product which they are also likely to buy.
Differentiating from Amazon
Even pure online players can provide things like advice via the phone – 15 minutes of talking to the customer can win a big sale. Even if Amazon had a call centre they couldn’t be expected to know in detail about every product they sell. If you are a specialised, top retailer you have that knowledge and so you can offer that interaction.
Online interaction with Amazon is not always the best. Our tool learns from all customers together, including both online and offline purchases. We determine how quickly customers use products and therefore when they might want to buy them again. Our Predictive Basket technology learns what people are buying on a regular basis and when so that online shopping for FMCG becomes faster than ever. With Amazon, they often suggest to me to buy things again even when it is something like an electric toothbrush that I won’t need more than one of.
AI – from hype to reality
Just because something has not worked in the past does not mean it won’t work now. Humans tried to fly for thousands of years without success but now for many people it feels like an everyday experience. A single small change, like James Watt increasing the efficiency of the steam engine from 0.5 percent to 3 percent, can make a technology viable. AI used to have almost no practical applications and neural networks needed teraflops of calculating power to gain limited results. But now AI hardware has become cheap enough to deploy, and deep learning has evolved, and thus using AI algorithms such as our Predictive Basket or Personalisation pays off.
The priority for retailers
Work on your strengths. Retailers need to differentiate where they can, copy where they can’t. Don’t try to innovate online if you aren’t an expert – just copy things from other successful retailers.
Strengthen your offline stores, provide great experiences there and have good people in your outlets – especially at your click and collect desks. Get those people away from unimportant tasks and put them in front of the customer.
Our targets for the next five years
We want to build up our European presence. We are a European company but we have not been in the UK that long and want to emphasise our pan-European credentials. We’ve also doubled headcount in our research department to build more great algorithms.
Our one line elevator pitch.
You have the visitors – we get them buying.
Cooksongold increases shop conversion rate by 26% with FACT-Finder
Cooksongold is a one-stop shop providing materials to jewellery-makers selling most of its 18,000 products through a 24/7 transactional website. This large and complex range can make it hard for customers to find the right products.
The company wanted to make it easier for customers to find what they are looking for through online search. It also wanted to create merchandised campaigns around sales events and drive new sales through product recommendations. With only a small marketing and IT team, Cooksongold needed to be able to do all this cost effectively and with minimal ongoing management. It also wanted to be able to replicate successful merchandised campaigns in the UK across several new markets in Europe.
Using FACT-Finder’s advanced filtering technology, Cooksongold can automatically display products based on their popularity or profitability. The retailer can also use FACT-Finder to create and deliver campaigns in half the time, allowing it to run many more highly profitable campaigns each year.
Cooksongold can also transfer its UK campaigns across other regions, with campaigns already having been replicated on local language sites for Spain and Germany. FACT-Finder automatically removes products that aren’t available in a particular country.
“We’re moving forward rapidly thanks to FACT-Finder’s unique capabilities. FACT-Finder has also provided exceptional support. Any issues we encounter are resolved rapidly. We also get analysis and regular recommendations on where we could improve,” says Jonathan Broadhurst, E-commerce Executive, Cooksongold.
Cooksongold’s IT team plans to take further advantage of FACT-Finder’s capabilities by improving personalisation. By doing so, customer experience, conversion rate and revenue will continue to improve. It also plans to use FACT-Finder across more European websites as it expands.
By making the most of FACT-Finder, Cooksongold has increased:
- E-commerce conversion rate by 26%
- Number of transactions by 26%
- Revenue by 22%
FACT-Finder IN BRIEF
Company Name: FACT-Finder (UK) Ltd.
Founder(s): Carsten Kraus
Founding Date: 2001
HQ Location: Cambridge
Size / Employees: 160
Customers: 1600 including Lidl and Footlocker
Carsten Kraus is founder of Omikron Data Quality, which offers the FACT-Finder product. His first major innovation as a teenager was a programming language that Atari bought and shipped with 700,000 computers. At an early age Carsten Kraus developed an understanding of and interest in logic and AI, going on to develop products which today provide master data management solutions for many large corporations and which are used by almost every European consumer when shopping online. Due to his expertise, he was elected Chairman of the AI Division at bwcon. He has applied for several patents on AI processes over in recent years and is a popular speaker at conferences.
This Company Spotlight was produced by InternetRetailing and paid for by FACT-Finder. Funding articles in this way allows us to explore topics and present relevant services and information that we believe our readers will find of interest.