Beat using big data: Koen Pauwels


Better use of your marketing resources to achieve company objectives. In boom times, that typically means faster and more profitable growth than the competition. In recession times, reallocation decisions require insights on where to cut and where to spend.

Regalix: Your book is about Smarter Marketing, what does this mean?
KP: Better use of your marketing resources to achieve company objectives. In boom times, that typically means faster and more profitable growth than the competition. In recession times, reallocation decisions require insights on where to cut and where to spend.

Regalix: How can marketers optimize the use of analytics? What are the key metrics that a marketer should track?
KP: It depends on your brand position and country. The MSI best paper on Customer road to purchase: what consumers know/think (awareness, consideration), feel (preference, brand love) and do (purchase, repeat purchase, WOM). Online metrics of reaction to paid, owned and earned media. Depends on country: I find that brand love has more sales effects in mature markets, while consideration has higher effect in emerging markets.

Regalix: Does analytics make marketers more accountable?
KP: Yes, if you use it well! It is a closed-loop system where you predict the effects of your proposed actions, decide among actions (partly) based on their predicted effects, execute and monitor the effects and then learn afterwards whether you should update your expectations. Finance executives are typically too prevention-focused, while marketing executives are too promotion-focused: they overestimate effects and don’t account for uncertainty and risk.

Regalix: Can it be misused to disguise poor performance indicators?
KP: It does depend on how you use it some metrics are indeed poor performance indicators, for instance the number of Facebook likes/followers and ‘talking about’ metrics. Indicators need to be connected to each other, to marketing and ultimately to business performance. Chapter 8 details how you can do this and thus move from tracking dozens of ‘KPIs’ (many of which are poor performance indicators or PPIs instead) to tracking the few key leading performance indicators (KLPI). For instance, we showed a large retailer that specific social media conversations (1) drive actual store traffic and sales and (3) are influenced by specific marketing actions.

Regalix: Does data analytics hamper a marketer’s creativity process?
KP: No, instead analytics help make good creative shine by showing how much it has contributed to the brand. In my experience, creative differences have huge effects on sales and brand equity performance. Before analytics and dashboards, you had to ask your client or boss to trust you that the creative idea drove sales. Now you can demonstrate it in a language they understand and accept. Of course there has to remain room for experimenting with ideas that don’t lead to measurable results…about 10% of the marketing budget should be set aside for those.

Regalix: What are the key benefits and challenges of using analytics?
KP: The key benefits are better decisions, higher profits, more learning and higher accountability and clarity. The key challenges are obtaining and maintaining high-quality data and showing decision makers how analytics is not replacing their intuition, but enriching it to the benefit of their organization and their own career.

Regalix: How does the application of analytics vary across different industries and stages of the buying cycle?
KP: Highly competitive industries see more use of analytics, because better decision making is a key differentiator when consumers are highly demanding and competitors have copied your product and services. These industries also see the highest Return on Assets for analytics use, increasing by 21% (versus 8% in general industries (see the paper by Germann and colleagues in the International Journal of Research in Marketing). As detailed in the book, I have personally helped companies across 3 continents apply analytics in industries ranging from fast moving consumer goods to durables (cars and electronics) to services (travel and banking) to business-to-business (telecom, office furniture).

Analytics can help throughout the buying cycle, focusing on converting prospects into category consumers in early stages (need recognition and awareness), on persuading first-time and competitors’ customers to buy your offer in the consideration and preference stages, and on conversing with and keeping your own customers in the experience and loyalty stages of the buying cycle.

Regalix: How can marketing analytics help create messages for our target audience?
KP: When preparing for message creation, analytics can help you get greater insight in not just what your audience wants but also how and when they want it, as evident in their online browsing behavior. After creating potential messages, analytics helps you to test audience response to messages in real-time, so you can adjust your message for best results. For instance, the Obama election campaign created several executions of its online donation campaign, varying both the visual (picture of Obama by himself, picture of his family, video) and the slogan (act now, ask for more information, etc). To their surprise, the campaign with the highest click-through combined the family picture (not the video) with the ‘ask for more information’ (not the immediate call to action).

Regalix: In analytics, how important is the process of testing?
KP: Testing is very important on several levels. First, you need to test your implicit model of how the world works and you need to test the analytics model to verify its performance across consumers, regions and time periods. Second, you need to test whether the analytics perform at a tactical level, e.g. selecting the best execution (as in the Obama example above). Finally, you need to test the strategic implications of analytics by reallocating your marketing budget and monitoring the results. For instance, the office furniture company cut by half its spending on the main marketing action (which analytics showed to be ineffective) and doubled it on its online paid search (which analytics showed to be very effective). To test this, we split up the country in 4 regions with similar past performance, and used 1 region as a control group (no change), changed only the main action in another region (cost cutting), doubled online search in the other, and implemented both in the 4th region. Net profits increased 14 times more in the 4th region versus the control region. (for details, please see the video at MIT).

Regalix: How can marketers enhance customer knowledge based on analytics?
KP: Customer knowledge comes in many sizes, and analytics can help you in each:
1) assess market size and potential by quantifying the perceived need for your product
2) get better insights into not just what your potential customers want but also how and when they want it, as evident in their online browsing behavior.
3) analyze consumer response to marketing actions, including how it makes them talk about it on social media, perform searches etc.

Regalix: In analytics, what are the key performance indicators?
KP: For analytics? Better decisions, improvements in company-level objectives (eg profits, market share or reach), efficiency (ROI). For the organization, it depends on the goal, eg increasing volunteering or money raised for NGOs, higher sales or stock market capitalization for others. Potential KPIs derive from these goals, as explained in chapter 7.

Regalix: What is the future of analytics and big data?
KP: It follows the Gartner hype cycle: we are now past the peak of expectations and will see the difference between organizations that have fully embraced it versus those that are paying superficial lip service to it. After the shake-out of consultants in the ‘through of disillusionment’, we reach the ‘plateau of productivity’ where analytics and big data are fully integrated in the way we do business.

Regalix: Can you share some examples of how companies have successfully used analytics?
KP: Yes, these are easily outlined as case studies in the book. Besides the one noted in question 8. I like the US branch of a foreign car manufacturer. Foreign headquarters slowed down product development from 4 to 5 years and cut advertising budgets, but still wanted US management to hit increasing market share and profit targets. We developed an analytical model explaining their sales by age of the fleet, price and incentives, offline and online advertising, distribution, product quality and competitive incentives. Next we gave decision makers a dashboard that allowed them to change the marketing budgets in slide bars and in real-time observe the projected profits. It clearly showed how the targets became unrealistic as the higher age of the fleet and lower ad budgets would substantially reduce sales. US management used this tool to negotiate with foreign headquarters and obtain better budget and allocation, raising profits and satisfaction on both sides of the ocean.


Interested in commissioning a study or other thought leadership?