How to Focus on “What Consumers Do, Not What They Say They Do”


We all like to share memories and tell stories about what we did during our last vacation, our last weekend, or our last night out. Unfortunately, as time passes, details fade away or become distorted. Sometimes, we even conflate our recent experiences with those from the past.

The brain is a fantastic machine, but it is not without faults-neither are our memories.  What we say we did is never entirely loyal to what we actually did. This phenomenon is true for everyone, including consumers.

Fáilte Ireland, the National Tourism Development Authority for Ireland, endeavored to maximize tourism revenue in rural areas, those places off the beaten path. They needed to learn what tourists did during their last vacations and identify potential rural hotspots.

Despite the more traditional approaches of market research to diminish memory biases, tourists can only tell you what they remember they did. How can we be sure these accounts are what really happened?

Your source must be factual and objective

What is one of the most factual and objective pieces of data you can find? Visual data or photos/videos. Why? Contrary to a memory, a photo doesn’t miss details, interpret or distort reality like a memory or a questionnaire could.

Visual data can show you exactly what happened, where and what types of emotions were displayed. Plus, it often displays secondary or “accidental” relevant information. Consumers will take a selfie to remember their trip to Ireland, but not to remember the shirt or sunglasses brand they wore. With a photo, there is the chance for additional data collection, such as the clothing choices of vacationers. Something they would probably not remember during an interview could still be valuable for an analyst.

Only after you observe this objective and factual data can you integrate consumers’ viewpoints to enrich your own perception.

Please note that we are dealing here with regular consumers, not professional bloggers or photographers with staged content.

Prefer an automated solution to manual coding

Finding a significant sample of consumer photos for a qualitative analysis -minimum of 500 based on our experience- is not difficult, especially with the success of visual social media platforms such as Instagram.

However, scoping, tagging, classifying, and analyzing these photos, unless you decide to study a small and potentially biased sample (e.g 50 photos), analysts might need several days to build a representative sample and extract insights. You need to have the right consumers, with the right geolocation and find a classification system that leaves no room for interpretation.  

An automated solution, such as Computer Vision, will deliver the same insights in hours and will integrate--in real-time-- any additional uploaded content.

Bridge Qualitative and Quantitative insights

The ideal solution to understanding consumer activity is to combine qualitative and quantitative analysis; observing thousands of photos from hundreds of consumers allows for unique insight into their thoughts and experiences. Where traditional approaches often cannot deliver these results due to budget or time constraints. Computer Vision can.

Analysis in action

In the case of Failte Ireland, we discovered (by analysis of over 36,000 photos) that among tourists visiting Ireland’s more rural areas. County Clare was the most photographed county, while County Galway was a close second, especially thanks to Galway city.

Galway city was pictured as a vibrant city, full of music “with great food” and “amazing people”. Its waterfront (Cladagh) heavily contributes to its success, being an iconic photographic place.

Find out more a lot more about tourist activities, profiles, journeys and most visited sites by into the case study



Just as in Failte Ireland, the combination of A.I. and Computer Vision can allow you to extract insights directly from what tourists did, without the memory biases that traditional approaches carry. If you are interested in knowing more about this type of approach, do not hesitate to contact us and request a demo.