I use data science
for customer insight

Have a Scroll

What's your edge?

Now that everyone has the same software, platforms and tools, advantage comes from what you do with them.

I will never forget the first time I made dry maths useful. For the past 20 years I've been doing exactly that, across retail, pharma, auto, media and tech, with celebrated brands and clever agencies.

My clients are progressive insight agencies looking to supercharge their analytics offering, and independent consultants winning technically challenging projects. They rely on me to challenge their brief, ask the difficult questions, lay bare the risks and deliver practical analytical solutions.

Accessibility

It has to be simple

When simplicity has traction, sophistication and nuance be damned.

Value

It has to be efficient

I use open source machine learning and automation to deliver cost effective analyses. My deliverables are defined in advance.

Validity

It has to be repeatable

Insight is not lucky - anything that cannot be repeated is not fit for purpose.

Impact

It has to look good

The packaging of insights, with the audience in mind, is the difference between success and failure.

Ryan is an astounding mix of technical brilliance, commercial acumen, and common sense. Not only has Ryan been able to deliver exactly what we (and our clients) need, he has also been able to add value through general observations, processes, and other IP - ensuring he leaves a legacy on our agency. A rarity in the industry, and thoroughly recommended.

Simon Thompson, MD, Relish

Ryan is extremely creative and versatile, he combines strong analytical and marketing science skills with an impressive ability for design and storytelling. His technical skills are exceptionally diverse... I would recommend Ryan to any company looking for an analytics leader with verve and imagination.

Frank Hedler, Vice President Data Science, GfK

I love working with Ryan. He can explain super complex data stuff so that even I understand it. And he's very good at making things look awesome. But not in a shallow way, in a way that's purposeful.

Alex Hodge, Founder, Yugen Research

You want Ryan on your team. Commercially focused and excellent at the technical stuff, Ryan has that rare talent of providing practical and effective solutions and doing it in such a way that everyone in the room understands (even me!). He is always pushing to innovate and move the thinking and science of data on, but equally kills at the old-school stats and data approaches. Perhaps most importantly, he is collaborative, enthusiastic and enjoyable to work with. Working his magic behind the scenes, or sitting down and presenting to a client, Ryan is the real deal.

Paul Griffiths, Evolve your Insights

Analytics.

Here are my favoured go-to techniques

Strategic Brand Building

Key Drivers, Penalty Reward and traditional SEM to create meaningful brand health KPIs, brand positioning and equity models

Get in touch

Purchase Journey

Marcov Chains to isolate triggers and bottlenecks across brand touchpoints

Get in touch

Customer Experience

Kano analysis to prioritise delighters and hygiene factors

Get in touch

Price Optimisation

Gabor Granger all the way through through monadic designs to derive price elasticities

Get in touch

Data Fusion

Classic, constrained & hotdecking to create synthentic datasets

Marketing Mix Modelling

Traditional data modelling of market share, media spend, web analytics and surveys to demontrate ROI by channel

Get in touch

Segmentation

Model based clustering techniques, most frequently latent class, to form meaningful and coherant customers groups

Get in touch

Text Mining

Word embedding & LDA for topic modelling and sentiment analysis

Get in touch

Time Series

Facebook's Prophet library with tradtional ARIMA to detect and predict trends

Get in touch

Probabilistic Inference

PyMC3 for A/B Testing and operational optimisation on the fly

Get in touch

Trade Off Designs

Max Diff, Choice based Conjoint and adaptive designs for new product development and range optimisation

Get in touch

Anomoly Detection

Isolation Forests and SVM for tagging erroneous transactions, interactions for sales opportunities, fraud detection and automated data cleaning

Get in touch

Why me?

I don't seek to add value with machine learning - it is a readily available open source technology that costs pennies to run. ML is a job for software and there are plenty of great options to choose from.

I add value by following the "Whys", bringing years of experience in the insights industry, domain knowledge and creativity.

My role is to leverage data science's unparralleled strengths and avoid its embarassing limitations as I repackage and translate the latest innovations into what makes sense for the insights industry.

Data crunching is one thing. Making data shine is another.

As for influencing minds with it? Well...

Shall we get started?

Views.

All Methodology Machine Learning Techniques