Adore Me was founded in 2012 by Morgan Hermand-Waiche, after struggling to find lingerie for his girlfriend which was affordable and high quality. His business has now raised $11.5 million in capital and employs 100 employees in the US and Europe. The e-tailer's revenue has tripled year-on-year, hitting $43 million in 2015. It is now the second-fastest growing retail company in the US, outselling Agent Provocateur and very quickly disrupting the lingerie market which has been dominated by Victoria's Secret.

Currently trading in the US and Canada, Adore Me's VP of business development, Sharon Klapka, shares her learnings about A/B testing in the retail industry.

Why is it important for eCommerce websites to be conducting A/B testing?

We're an online brand, 100% of our data comes from online, while only 10% of Victoria's Secret sales come from online, so they have limited access to data. We really have an amazing tool which if we don't leverage we are losing so much. That's a huge advantage when you're a pureplay eCommerce, we know everything, and because we're a start-up, we need to make the most of our marketing dollars. And A/B testing meant we really tapped into making our marketing amazing, which really helped us grow so quickly.

Because we're fast fashion and launch a new collection every month with 30-40 items, we don't want people to drown in a sea lingerie, so we ask customers to fill in a short quiz when they visit our website with their preferences. They then receive a curated shop with the items we think are going to suit them.

The fact we can A/B test and see what people are really attracted to, what they're engaged with, what resonates with them, that enables us to create a far more personalised experience. For instance I know if we use a plus-sized model, even petite customers relate to her and it creates a stronger emotional connection – there's this assumption that I'm always going to want to engage with size zero, super glamazon, with a six pack.

The model makes a huge difference – we can have two identical creatives, same copy, photographer, background, same everything, but people can engage with a certain model and not another. Through A/B testing we found out blondes don't sell and people engage more with brunettes, the other thing we saw was customers engage more with models who have their hand in their hair, versus their hand on their hip, which is ridiculous and of course we can't have 90% of the models on our website with their hand in their hair, but by far it's the best performing images and we know we're not optimising the amount of sales when we switch it.

What advice would you give retailers about to embark on agile testing?

At the end of the day, you have to find the right people to do the A/B testing. Companies with more traditional backgrounds that don't come from the online realm, they get really flustered and go and interact with third parties which don't always provide the tools. So firstly, do you research, see who can provide you with the tools to do effective A/B testing. And for us, it was always doing it in house. We're not big fans of third party.

We found with A/B testing some changes resulted in 340% more sales, these stats make a huge difference, but you have to be resilient and continue testing, you can't just do it for one month.

The biggest comment I have about A/B testing is use the insights, because I hear a lot of brands –surprisingly big online, pureplay eComm brands – see their A/B results but they don't act on them, because 'we've been in fashion for 15 years and we know what is right'. They're looking at the data, but still insist they are right. I think that is the biggest mistake brands can do – you have the data in front of your face, but you still think that you know better.

How do you find the digital talent?

If you go to Sillicon Valley you have to pay a preposterous amount, and when you start a business in New York you can pay them less preposterous salaries, but they're still preposterous, which is why our IT team is in Romania. They are extremely talented people and their CIO was out of IBM, so in that sense we can have extremely skilled people and it is not going to require us to raise $100 million.

We have people with MBAs and those who worked at McKinsey who could have gone to Amazon, but really it was the fact that we have a really disruptive vision. We're aiming to disrupt the entire market and we're very open about it, I think that just attracts amazing talent. And we're a very international team, everyone is young, hungry, driven and that attracts a lot of people.

What can retailers learn from start-ups?

The first tip is the culture – you need to attract the right people. For example 80% of our traffic comes in from mobile, we're advertising on TV, but we have Instagram and Snapchat and you need young people to do Snapchat – I don't have Snapchat. Make sure you attract people who are creative, hungry, young, know what is next, and aren't afraid to say their mind.

Also, don't think you know better than the data. The data isn't wrong, you are wrong, you need to adapt. There's a huge shift retailers need to do.

Finally, don't assume everything that has been working for you in the last 20 years is going to work for you in the next 20. I think that's a huge mistake you see in a lot of markets which have been disrupted, that behemoths multi-billion dollar companies didn't realise in time that change was happening.