Changing marketplace and societal dynamics

8 mistakes retailers are making with personalization

By Frédéric Thomas-Dupuis, Randall Sargent, and Danny Zhang

Customers are increasingly expecting personalized recommendations, offers, and greetings through a seamless omnichannel experience. In fact, according to Salesforce, 73% of consumers expect brands to understand their unique needs and expectations, while 56% expect all offers to be personalized. Globally, 70% of business leaders are boosting investments in personalization, business intelligence platform Statista finds.

However, while many retailers have only recently developed personalization capabilities, and few have truly mastered them, we can learn from the common mistakes observed in the industry.

1. Start with coordinated vision

Personalization can mean different things to different people, from personally greeting top customers to personalizing search results on a retailer’s webpage to serving up the next best-personalized offer. Without a coordinated vision, efforts can appear disjointed to customers who receive mixed messaging and overlapping deals. As retailers embark on building personalization capabilities, it’s important to build consensus around a clear destination so various teams can work towards the same goal.

2. Invest in yout loyalty program in parallel

Personalization requires rich customer loyalty data to make recommendations most relevant. Retailers should focus on driving loyalty sign-ups through in-store signage (for example, QR codes) and campaigns (for example, extra bonus points for creating a profile or downloading the app) in parallel to or in advance of building personalization. In a recent pilot, we found that a combination of personalized communication and in-store signage and kiosks resulted in 14% more new sign-ups. Personalized offers and communication can help drive engagement, but retailers need a strong foundation of customers they can track and reach to reap the full benefits of personalization.

3. Personalization is not just advanced targeting or segmentation

True one-to-one personalization requires a complete flip in mindset from the traditional ways of marketing, where a company develops an offer or campaign and then determines the right customer segment to send it to. A better approach is to start with customers and determine through a combination of business rules (for example, what are the priorities and strategic decisions to layer in) and algorithms (for example, collaborative filtering, and propensity modeling) what each should get from a bank of content, offers, and campaigns.

4. Off-the-shelf technologies are (often) necessary but not sufficient

There is no single solution that comprehensively delivers one-to-one personalization. Instead, building a personalization capability requires stitching and tailoring various types of technology together: New machine learning/artificial intelligence algorithms (“the brains”) should be built in a flexible and scalable environment and delivered to customers via existing off-the-shelf solutions. In a recent pilot, we leveraged the retailer’s CRM software to deliver personalization to customers, feeding it the outputs of algorithms we built that flexibly ranked the most relevant offers for each customer.

Picture of an employee making a sandwich in a store

5. Curating existing offers drivers significant value

Many retailers feel that personalization requires multiplying their volume of content, offers, and campaigns. In reality, they can drive substantial financial impact through curation of what they already have. For example, a South American grocer recently launched a personalized digital flyer that led to a 2%-5% increase in customer spend, simply by curating existing offers and content without introducing any new personalized offers or discounts.

6. Fenerative artificial intelligence (AI) can help, now

Generative AI can help to increase the amount of content, offers, and campaigns without requiring significant additional creative resources. We’ve tested leveraging generative AI to personalize email greetings to customers, finding that it drives twice as many click-through rates. This new technology requires human oversight and review but can massively accelerate creation.

7. Designate a cross-functional team

Personalization requires coordination across many parts of the organization, including digital, marketing, merchandising, and IT. We see the most effective organizations creating a small, dedicated cross-functional team focused solely on building and sustaining personalization capabilities and coaching and challenging the rest of the organization to make use of what’s available.

8. Take action before external disruption forces your hand

In jurisdictions around the world, local governments are starting to limit print flyer distribution or ban it altogether for environmental reasons. In other markets, sudden increases in printing and distribution costs can leave retailers stranded. Creating a digital alternative that can be personalized will allow retailers to stay on the front foot.

The rapid advance of technology has opened the doors for exciting personalization opportunities. As retailers develop these capabilities and unveil this technology, it will be important to keep these lessons top of mind.