Tag Archives: Insights Association

Insights Association Paradigm Shift: Chris Whitaker, Explorer Research; Steve Olsen, Nestle; Maja Neale, BMO

My notes. Any errors are my own.


Better Behaviour Predictions Through Better Behaviour Testing by Chris Whitaker, Explorer Research

Traditional research plus non-conscious research = richer insights

People don’t automatically remember the details of every shopping trip or why they made their decisions.

Recall is not a reliable measure of real shopping.

When we get a new job, we plan the entire route as soon as we can. But a year later, we just don’t think about it. We can arrive without even realizing how we got there. Planning takes up a lot of energy and time so we remove that over time. Decisions must become easier over time and this comes from using the unconscious mind.

People inflate recall, purchase.

Only 60% of brand purchases are recalled correctly. Eye-tracking knows precisely what you bought.

Only 48% of people recall the flavour or variety purchased. Only 27% recall the size correctly. Dollars spent are recalled 50% higher. Time spent is remembered as twice as long as actual. Brand noticed first is only correctly answered by 4% of people. Eye-tracking tells you exactly what they looked at first. And what was second and how long for each one.

Context influences behaviour.

What happens when you buy a whole case of wine because it’s the best wine ever but then you discover it’s not as great as what you thought when you first tried it with your friends at a beautiful winery.

TV ads are recalled a bit more than online ads. Purchase intent is about the same. However, biometrics are quite different. Online is 30% more engaging. Engagement plus distractions is three times better for online ads.

People say they seek price first but we rarely go through a list of needs, we just buy what we always buy.

Consumers idealize their behaviours.

We need to use behavioural measurements. VR, shopper labs, AI, eye-tracking, facial coding, mobile measurement.

In situation testing is more predictive and it can be timely and cost effective.

Yesterday and Today – The Nestle Canada Story by Steve Olsen, Nestle

Appreciating nuances vs clinging to arbitrary benchmarks. A slightly lower score isn’t “didn’t hit the hurdle,” it’s a discussion point.

We need to value feelings as much as facts. Get in front of consumers and talk to them, bring the experience to life, experience reality.

What we can do differently

Be more collaborative and less transactional with research partners. Together, you can avoid rabbit holes. It’s okay to show things that aren’t fully finished. This ensures you get to the right solution more quickly and easily.

Try new tools and techniques

Think about discrete choice, implicit testing. Remember that the most a consumer has thought about your product is probably during the minute you mention the name to them.

Prioritize foundational research

U&As can be used multiple times over the years and when combined with new information can still reveal insights.

Expand expertise and influence beyond just research

Build a culture of learning throughout the organization. Become better influencers and presenters.

Continuous learning and inspiration

Go to conferences outside your specific area. Go to an academic conference, a marketing conference, design conferences. It can inspire you in your own work.

Suppliers need deeper business understanding. Be curious for context. Seek senior level face time. Be able to synthesize and integrate knowledge from other industries.

Develop interdisciplinary expertise, especially the social sciences. But recognize when you’re not the expert or when your tools aren’t as precise as you want them to be or when they can’t do what you want them to do.

Develop new technology capabilities. Consider user design and interfacing. Consider automation with customization. Figure out if you can scale what was previously unscalable, e.g., text analytics, picture classification.

Experiment with data and methods. Helpful for thought leadership speaking at conferences. Helpful for turning trials into subscriptions. Quantify the risk for potential clients.

Massive Marketing shift – How BMO Created a More Data Driven Business by Maja Neale, BMO

Brand challenges

Changing demographics, personal and privacy paradox, new media channels, interruption marketing. We need to maintain relevancy, measure investments, integrate across all channels, and rethink the way we market.

Banks have a responsibility to respect consumer data. But consumers expect what they receive to be tailored.

Trends in insights gathering.

Tech-based companies are growing. Self-serve DIY solutions are being rapidly adopted. Data integration across behavioural and company data is increasing. Demand to quantity financial value of insights is increasing.

Most companies sit between analysis and predictive rather than applying predictive knowledge to dictate behaviours necessary now.

Try to create micro-level results, e.g., tell each bank branch precisely how to improve their branch. There is sufficient transactional data to make this happen.

DO SOMETHING is the key important take away for any research. Once you have the data, act on it with the aim to improve.

Integration, automation, and real-time are key to using and applying insights.

Insights Association Paradigm Shift: James Lachno and Nicholas Boles, Edelman; Hilary Borndahl, Kantar, and Nick Necsulescu, World Vision Canada; Ian Ash, Dig Insights Inc.


My notes. Any errors are my own. Make sure to read Ian’s at the bottom, nicely insightful.

Why Digital Data is the Fuel you Need to Make Fire Content, By James Lachno and Nicholas Boles, Edelman

74% of people avoid advertising.

Relatability is twice as important as popularity.

Audiences are now in control [i use multiple ad blockers, VPNs, and privacy browsers]

We need to always be relevant and valuable. We need channel agnostic stories. We need to be creative at the core.

The Story Planning Cycle

Identify your three or four key audiences. “Beats” as you might call them in journalism. Create a brief for each audience. What do people search for when they’re part of one of these audiences. What are the search and news trends on a seasonal level for these audiences. Conduct a collaborative content plan to be focused on the needs that are relevant to each audience. Build content, and decide on who, what, when, where to publish. Measure when people actually search for the various types of content to identify earned, organic engagement.

Going Viral Using Real Time Data

Need to recognize when a real time event could become a viral sensation that is in line with the business. E.g., #DartGuy smoking cessation.


  • Bake data mining in your strategic processes.
  • Collaboration is King
  • Bring clients on the journey with you
  • Speed is better than perfection [oh, remember the last speaker said that too!]
  • Don’t self censor

Unlocking ROI Through Advancements in Analytics, By Hilary Borndahl, Kantar, and Nick Necsulescu, World Vision Canada

Started the presentation by having audience log into menti.com: 42 57 80, a live voting website used throughout the presentation. [where I voted for “cat” about 300 times. Really nice live charts and word clouds based on our inputs. But it does draw attention away from the presentation.]

Marketers want balanced measurement but skew towards short-term tactics. Many feel they are leaving opportunity on the table.

Channel strategy and media tactics need to be tailored to address business objectives.

Drivers of ROI on TV media: Creative is huge, followed by market buy.

Niche brands are taking over the jobs that used to be done by key players.

Case study

2.3 million conversations evaluated. Ordered the data by brands, products, lifestyle, benefits, product features, and more

Brand by location was of key importance. Correspondence analysis identified opportunity gaps in the context of trends. They built a creative content strategy around these moments in time – celebration and togetherness.

Machine learning and AI need to be integrated earlier into the process. Look at data second by second to predict branding, impact, facial coding emotion.

World Vision tested 14 creatives using AI/machine learning. AI traces KPI scores over a video duration to see impact of each creative. The model shows which ads have impact, generate spikes, where the spikes occur. Included a few ads that have been previously evaluated in a traditional form. Results correlated highly with models. Awareness was the only metric that did not correlate.

Love Can Do Anything was the winning ad.


Stuck in the Middle with You, by Ian Ash, Dig Insights Inc.

“Feel free to disagree with me or be offended.” 🙂

The world used to look like this: Automated/DIY, mainstream MR, high end custom, analytics/big data, strategy consultants.

Now the middle is squeezed: automated DIY and strategy consultants are squashing out the rest.

Strategy companies are on a buying spree. Automated companies are getting all the funding – Zappi, voxpop, qualtrics, surveymonkey

Private equity is throwing money at scalable companies.

SMall companies are competitive with companies that don’t have to generate profit – they just have to have a path to profit. So the small/medium companies suffer.

More consolidation is on the way.

Expect mainstream MR to buy automated/DIY companies.

Automated DIY will compete more directly with mainstream MR.

End clients are starting to buy analytics firms – McDonalds, Visa, Unilever have all bought analytics firms.

AI is the seasoning salt of software. Everyone says they have it, most actually do not.

AI is for prediction models, determining sample sizes, codon open end responses, statistical analyses, finding insights.

Merged data creates holistic approaches. Including sales and volition perditions, segment prediction models, price elasticity models, product optimization.

Blockchain allows buyers to interact directly with individuals. Protests privacy, incentives accuracy, creates fair compensation, provides data usage and payment transparency.

How to compete.

We all need to have automated methods. Develop data science capabilities. Increase customer service. Partner with digital ad agencies. Used merged data approaches with API integrations.

Insights Association Paradigm Shift: Ashik Bhat, Labatt Breweries of Canada; Andrew Go, Home Depot


My notes. Any errors are my own.

Sales Force of the Future

By Ashik Bhat, Labatt Breweries of Canada

  • The goal is for 80% of the plan to be perfect and then move at 100% speed
  • In the beer industry, any brand can replace another brand. You need your brand to be irreplaceable.
  • You need to understand the role your brand plays in consumers’ lives.
  • How do you become indispensable? Understand retailer challenges and partner with them to meet those needs. Understand consumer challenges with the beer category. Understand the sales team’s challenges.
  • Retailers: Retailers want to drive traffic or put more people in the seats.Alternatively, have people buy a more expensive product.
  • Consumers: Beer category is cluttered and disorganized. We must help consumers navigate this category. We need to make the choice easy. We need to educate people about beer, pairings, history without adding time to their day.
  • Sales team: They need to be more efficient. They need insights created in a customized way for their hyper-location, their market.
    • Using AI to drive account level assortment insights. Used stats can data, transactional data, sales output data, correlated with hyper-local geography. This data is updated weekly with new sales data. This helped retailers optimize their SKU and brand assortment which drives volume and incrementality. They can trade brands in or out appropriate. It also drives traffic. Ensures the right brands are listed in the right places – premium beers in the right places, light beers in the right places. And, this saved time. Data is fast, efficient, and relevant.
    • Sales grew 0.5% in a flat business and net promoter score is 9.3.
  • Consumer challenges
    • In other example, the retailers struggled to convert traffic into basket building. Conducted shop-alongs and learned navigation was a problem. Pricing wasn’t clear. Products were ordered by SKU or type, not brand. As a result, they implemented brand blocking. Also, people wanted cold beer but they didn’t want to enter a cold room. This problem was fixed with refrigeration. Finally, they ensured prices were clear fro every product. They added a blackboard that was colour coded with information about the beer so people were more knowledgeable.
    • Drove punches 8% and premium segment grew 1.7%. Shopping experience improved 50%.
  • Conclusions
    • Be business leaders, not just consumer experts.
    • Think about the customer AND the consumer: the retailer/store and the consumer. Both have challenges to face and solve.
    • Democratize data and insights so they are easily digestible. 15 graphs are not digestible. Be a story teller, not a data described.
    • Push past delivering research and focus on driving action through insights and data.

    One Home Depot: One Team, One Dream by Andrew Go, Home Depot

    • Aspiration is to be number one and most trusted home renovation store in the world.
    • People regularly shop on the website first before going to the store. The website is not for e-commerce, but rather the first part of the customer journey, to inform customers about what they might need to buy.
    • Analytics begin with raw unprocessed data, then structuring data into reporting, then contextualizing information to support insights, then tools to apply those insights.
    • Used three years of data to conducting basket analysis. Created a self-serve dashboard and insights were applied to assortment planning, planogramming, and marketing. Then recommendation algorithms. This allows you to stitch together all different aspects of the shopping journey into the holistic unified journey. Being PIPEDA compliant all the way.
    • This allows product classifications, supports anomaly detection, and supports competitive pricing. Need to make real time decisions to detect pricing errors, underpriced items, and overpriced items.
    • The data helps with onsite search. The website needs to learn the names used by consumers not names used by industry experts. It also helps with product rankings – filters need to be appropriate for category, for customer preferences. Need to take price and speed off the consideration set in the fulfillment process.
    • Marketing messaging also benefits. If they know you’ve bought a BBQ, they can offer BBQ covers or tools rather than promoting another BBQ.
    • Set up cues based on geography and weather – e.g., snow blowers need to be advertised in fall not in the middle of winter.
    • There are new data sources everyday that need to be unlocked.
    • Start with the “Why.” Focus on insights, solving problems for customers. Be sure to collaborate and avoid silos.
    %d bloggers like this: