CRAP Talks #11 – The TakeCROver

Another year, another CRAP, and for the first of 2019, we returned to the MOO Offices for the fifth time since the birth of CRAP. For I, and I hope I speak for many other long-time attendees, it felt nostalgic to be back in one of the original venues, with a TakeCROver from a variety of speakers, rather than the brand dominated talks of late.

As large and expansive as it is in terms of layout, the event space felt strangely intimate despite all the additional chairs we needed to set out aside of the designated seating area. It felt as though word had spread at the last minute and we were welcomingly over capacity.

Shahina Meru – “Unleashing The Power Of Pyschology For CRO”

First up was Shahina Meru, a Senior CRO Manager at Periscopix whose talk was entitled ‘Unleashing the Power of Psychology for CRO’.  She offered relatable insight into the art and science at play when a user decides to make a purchase and the different types of cognitive bias they experience.

She opened with an interactive experiment to prove her point, involving the two sides of the room being caught in a trick about the amount of litres in a 747 jet engine. By handing out opposing slips of paper to different sections of the audience, she was able to demonstrate how simple it is to be influenced by information revealed to the user ahead of their decision making.

It made me question how often I make illogical, irrational mental shortcuts to enable quick decisions, and when think-feel-do subconsciously becomes feel-do-think. She gave examples of anchoring techniques, the peak end and the bandwagon effect in the work ecommerce website do to attract more users and influence their decisions.

She gave outrageous examples of dark patterns which businesses have used, such as one company in particular who booked in a compulsory subscription renewal to the inbox calendars of their users. She concluded by saying that CRO is about people, and if you fail to respect those people then they will know they are being tricked.

Cyrus Azami – “CRO Challenges”

Cyrus Azami followed her talk – a Senior Digital Optimisation Analyst at MOO, who spoke about CRO challenges and the ways he chose to solve them. For him, at the heart of CRO is core problem solving which is crucial for optimizing your website. However, rather than looking to your customers to improve your CRO process, it is best to start from within and treat your CRO like your website, and your stakeholders like your customers.

He recalled his own experience of complexities in how he had to adapt to different industries, different goals and different stakeholders to make the goal of the CRO programme he was trying to achieve more successful. He learnt very quickly that not all CRO solutions are made equal, so you can’t replicate them and expect them to work in alternative scenarios.

He spoke about how both weak and strong ideas can lead to a disjointed approach, by either taking a random chance without any data in the former or by executing only one experiment on what you believe to be a finished article in the latter. He echoed Shahina’s point about the dark patterns and the motivations behind this type of behavior being very much lead by a fixation on business metrics.

To bring analysis back into the spotlight, he said misapplied data or simply adjusting it to adhere to intrinsic bias will not get you anywhere. So-called ‘ostrich CRO’ will only lead to delays on the results you’re hoping for and using data to effect can improve politically driven prioritization arguments in your CRO process. Managing expectations and educating stakeholders on what to expect is the way to go when achieving success in the long journey of perfecting and constantly improving your CRO programme.

Jonathan Brooks-Bartlett – “How Long Until We Can Roll Out This New Feature?”

Jonathan Brooks Bartlett, a Senior Data Scientist at Deliveroo gave us an inspiring introduction to power analysis and how sample sizes influence experimentation and A/B tests.

Not only was he quantitatively literate, but also an incredibly captivating speaker, with a superb sense of humour and an excellent knowledge of his topic. He let us know that even everything in a data science team is run through an experiment, relating to machine learning and new algorithms to predict customer churn.

He explained the importance of statistical power in data science and how this type of analysis simply allowed for a probability that will mean they correctly concluded that a new feature would be worth rolling out for a better result. The larger the sample size, the more equal the true value, and he evidenced this in a coin toss example to show how the frequency of a result can build confidence in telling the difference, even when you think it will be equal.

He gave a visual representation of critical errors and type I errors on a graph to indicate what is visible when predicting the success of the new feature and all the different elements considered before rolling it out. It was during this section when for the first time in my career was I told in the clearest way what a Monte Carlo simulation meant in relation to experiments.

He was able to share both his slides and blog links from a few of his colleagues at Deliveroo who have written entire posts on the subject, approaching a variety of problems with code. I found him to be a humble speaker but one who was one of the brightest and best at communicating complex technical topics to the audience with a sense of personality and charm.

Robin Mesnage – “How DNA Sequencing Is Changing Our Society”

Finally, I had the pleasure of introducing our non-industry related speaker, Robin Mesnage, a Computational Toxicologist at Kings College London who kept on trend by presenting his eye-opening research on how CRAP sequencing is changing our society. He began by explaining what DNA is, how and why they scientists sequence it and what we’ve learnt in the time this has been available.

He used a poetic metaphor to describe DNA sequencing as a book of life – a biography if you will – which can tell you about yourself and your health. The sentences are the genes and the letters are the composition of those. Discovering this was the greatest challenge in scientific history, and thus, despite cost and amount of time needed, making the gigantic puzzle to rebuild a human genome.

This sequencing does end at humans either – they also retrieved samples from prehistoric animals from over seven hundred thousand years ago and reconstituted an original 5-frame gif from DNA which became a huge milestone for the storage of information. Multimedia aside, the more useful way for DNA sequencing to have an effect on human life is through genomic medicine.

Cures for cancer can be result in higher success rates by sequencing, identifying and strategizing for personalized array chemotherapy. He also spoke about consumer genetic testing, and why these results shouldn’t be used in clinical researches, simply because of the amount of false positives in developing diseases because of your DNA using purely this method. It’s nowhere near as sophisticated, and disregards the strong impact of environment on our health.

He reminded us that despite the connotations of bacteria being seen as a harmful substance, the ones we have on our skin and in our bodies are actually highly protective. Once the gut microbiome was cultured and sequenced, it revealed its importance in controlling diseases such as diabetes, autism and obesity. There has been a crap-sule (or fecal transplant) created which is being trialled currently that people can take to change their gut microbiome to that of a healthy person in order to lose weight This, of course, still carries its risks.

He focused the last section of his talk on his particular study into the glyphosate pesticide’s effect on the efficiency of modern medicine as well as how it’s related to certain types of cancer. We currently can’t escape it – it’s present even on organic foods, and if everyone’s urine was tested in the room then it would contain certain levels of this dangerous pesticide, which is meant to have no effect on human cells.

Despite being a heavy topic, Robin guided us through with humour along the way and came armed with live demos to share with the audience during the networking time which was very entertaining (don’t worry, he didn’t have any samples with him – it came in the form of a portable DNA sequencing machine!). The non-industry speakers, as a staple of both the brand takeovers and the usual format, never fail to inspire and tonight was no exception.

There was an amazing energy of excitement in the room – one which showed how eager everyone was to learn and jump into 2019. After the success of the previous diversities CRAP has shown, it feels as though it is continuing to shake up the events space and will continue to appeal to more sub-related specialisms as it grows.