Recommendations Must Be Thoughtful, Not Random

When content or product recommendations miss the mark, they can go from valuable to annoying (or offensive) in less time than it takes to X them out. Here's how to avoid that.

When content or product recommendations miss the mark, they can go from valuable to annoying in less time than it takes to X them out.

Alhough social networks, native networks and marketers know that relevant recommendations are important, most have yet to master the art. And when recommendations are sloppy and don’t work, they deflate the interests of consumers, and can even offend them.

The unexpected consumer

Consider: Demographic data might lead you to assume that every stay-at-home mom wants to be inundated with recommendations for cooking and cleaning products, and every senior citizen can benefit from countless ads for incontinence products, but of course that’s not true. Buying into that kind of assumption ignores the reality that individuals have individual needs, tastes and preferences.

For recommendations to work, they have to closely align with individual consumers’ interests. And assumptions will drive potential consumers away. Online is noisy enough without you adding to it with spammy offerings that aren’t relevant.

Analyzing conversations on social media can demystify consumer desires, especially if you mine for psychographic insights around emotions and behaviors. It’s not simple to do, but it is becoming essential, allowing you to see the “feelings, attitudes and commonalities across multiple demographics” and uncover target audiences you never knew existed. Going after your expected audience is so 2015.

Quality matters

Recommendations must also demonstrate respect for consumers’ time and intelligence. That means they should be high-quality and relevant—not spammy clickbait. As Revcontent advises, “One of the fastest ways to drive your audience away is low-quality content recommendations.”

Sponsored and recommended content must be on par with your own, both in terms of relevance to your field of expertise and quality of presentation. Bad content at the bottom of your blog doesn’t make yours look better—it makes your brand overall look worse. At the least, it leaves visitors questioning your taste; at worst it makes them wonder if your product or service is just as low-quality.

And it’s the same with product recommendations. If a consumer purchases a luxury item, the recommendations that appear post-purchase shouldn’t be for items they’d find at the local dollar store. It’s not about making a recommendation for the sake of it—it’s about making the right recommendation for each user.

Hone in on your target with Facebook ads

If your business is on Facebook, you have a number of options to ensure your recommendations are hitting their mark.

A good place to start is with your current followers. Each Facebook user who likes your brand’s page has a whole network of potentially interested friends you can target. Like minds attract one another, so chances are whatever your current users find appealing will also be attractive to many in their network of friends.

Facebook also lets you target consumers by their interests to define your brand’s ideal audience. Through likes, shared content and keywords, Facebook users give you a lot of information about what’s important to them. By integrating this data into your recommendations, you can give users exactly what they’re looking for, compelling them to connect with you in turn.

Tip: Keep in mind that while you want your target audience to be specific, you don’t want to alienate potential consumers by making it too small. Over-refine and you run the risk of cutting out a segment likely to respond positively to your brand. 

When in doubt, test!

If you want to dial things up yet another notch, A/B testing on Facebook lets you assess and perfect the effectiveness of your message through side-by-side comparisons. To get the most out of your recommendations, you’ll want to put a few variations out there at first to see what users respond to.