Kindness is at the heart of the recommendation. It’s one person saying to another, I see your situation and I think this thing can help you. That gets lost in an era of mass, computerised (and often low-quality) recommendations. But I’m talking about why, really, recommendations (and the members of the genus: suggestions, tips, hacks, help, advice, counsel) are offered in the first place. We’re able to make a good recommendation when:
- we perceive a need
- we know about something (or someone) that can help address that need
- the beneficiary doesn’t have that knowledge.
Reminding us of that, helps with the next question:
What makes a good or b
There are many examples of terrible recommendations. You probably have several in your inbox right now. The LinkedIn algorithms make them often. Even the likes of Twitter, Amazon and Netflix often make mediocre recommendations, dismissed by users without thought. In learning, the recommendations we get are worse still and for the most part devoid of love, kindness or intelligence. Made with minimal information about the user, they tend to be unspecific and unexplained; the same recommendations could have been (and probably were) sent to many other people in exactly this form.
Here’s a better recommendation. A couple of weeks ago, I was talking to a colleague. We were talking about travel vs settling down and he told me he wanted to buy a place and that he and his fiancee had just seen one at the weekend which they loved...but should they make an offer on the first place they saw? This reminded me of a problem described in a book I read last year. The book’s called Algorithms to Live By and the problem’s called the Optimal Stopping Problem. This is essentially about how to use maths to decide when to ‘settle’, be it for flat-to-rent, house-to-buy, parking space, life partner, etc. For me this is a perfect example of a good recommendation: I was able to perceive a need and knew something that might help address that need, a connection the beneficiaries weren’t able to make .
Breaking it down
Over the past few months, I’ve become especially interested in the links we as a company share (mostly on Slack and email). We’re very good at finding stuff out. We’re also pretty good at sharing it. But we can be a lot better at making the recommendation effective (and kind). A lot of our shares are literally just a pasted url. That, of course, is vital. But more helpful are these four contextualising Ws:
- What it is, summarised. A salient excerpt may be even more effective.
- Who, specifically, might find it interesting. Probably not everyone will.
- Why might it be interesting for that person / those people (and the more detail the better, not just at job role level but because you happen to know Cindy’s looking for video editing software that might do X, or Naomi’s writing an article about Y)
- And all of these contribute to a fourth factor - from Whom the recommendation comes. If you rate the recommender, you’re more likely to take the recommendation.
Note that these require the recommender to have engaged with the content him/herself. Sadly, ludicrously, often absent.
Recommendations made with love, care and kindness (ie the 4 Ws above) get more engagement and do more net good (despite taking longer to post) than just pasted urls to everyone. I’d much rather make fewer recommendations which made an impact than lots which are largely ignored. You’ve probably (subconsciously) noticed this already; the LinkedIn reposts with no contextualising commentary rarely attract much attention in the form of Likes, Comments or clicks. Though it’s easy to fire off a recommendation in seconds these days, a good recommendation takes a few minutes to make; it’s an artisanal activity.
Love, care and kindness is one - fanciful perhaps - way of looking at recommendations. Another is just that they’re personalized, ie designed to meet someone’s individual requirements. Recommendations that are carefully personalized in order to be useful are more likely to be acted on. This word - personalized - has been tainted in recent years with the broad brush of technology and mass communications. This is a result of high-quantity, bad-quality personalization, not personalization per se. Intelligent, kind personalization (mostly from person to person) is still beautiful. We must do as much as is humanlypossible to get the right stuff to the right people.
Get better at it, help people more. For us, our job is to bottle that human thinking process and emulate it through technology, which includes our AI stack, magpie. And this is a quintessential artificial intelligence project: building technology to learn from, copy and try to improve on human...recommendations.
Find other articles like this
This post is part of a series that my colleagues and I at Filtered are working on. It’s broadly about how recommendations help us to make sense of all the content clutter, especially in learning. Have a look at our other posts here.
New L&D community
We’ve recently launched a free new community for L&D professionals on Slack. Every Tuesday we’ll be holding live discussions at 4 PM GMT on topics including virtual classrooms, content curation, artificial intelligence and more.
Intelligent learning recommendations
Last year we launched a new tool to provide personalized learning experiences for L&D and HR professionals called globalfilter for L&D. It's an online recommendation engine with over 200 high-quality learning assets to read, watch, practice and apply for our industry. Click on the link above to try it out, or trial our version with 500+ assets for all professions here.