How I used LinkedIn polls to find qualified people for free.

LinkedIn polls were an amazing way to build an audience that explicitly identified potential behaviours relevant for Knowcast’s user needs research.

In my previous post I shared why I chose to invest heavily in behaviour-based research to understand Knowcast’s audience and why I chose a diary study to do this.

Running the diary study

Running the study involved the following steps, with three separate two-week cohorts, with 12 participants in total:

• Plan: what am I going to do?
• Execute: do the actual study
• Analyse: look at results and for insights to help crystallise the nature of the audience, the problems they have, when, where, how, what they do now, and ideally some sense of cost they will pay to solve it.

I break down the diary study planning into three steps:

• Define the audience and participants
• Define the method
• Select the supporting tools

Searching for the Knowcast audience

Firstly I talked to a couple of agencies who run diary studies for user research. What I learned was:

• The agencies usually research general consumer behaviour such as supermarket purchases. I was warned that my audience “well-educated affluent knowledge professionals” will be hard to find because they are very expensive and value their time more than is typically recompensed for 3 hours of contribution (approx £50).
• Diary studies usually span “a few weeks” with two weeks being pretty reasonable but can go on as long as six weeks.
• The number of people varies a whole bunch (between 4 and 20) but one agency suggested “15 and expect about 20% or 3 people to drop out”.
• Agencies are expensive: in the UK, hiring an agency to run the diary study ran to around £500 per participant. This included participant sourcing, agency project management of execution and report preparation. These numbers were very useful as an upper bound for what I needed to pay. Could I do better?

In the end I decided to use my own network and run the research myself. I had an upper bound of £500 per participant, so any less than that was a bonus.

“Scratch your own itch” comes into its own again

As I’m trying to solve a problem I have personally, there’s a good chance there are people like me in my network. So I used my LinkedIn profile (around 9k connections), plus direct outreach into a network called The Founders Network ($500 or so a year I think when I paid, looks like it’s $999/year now), and a few others. LinkedIn was the most cost-effective, being free + 15 years of network building.

A research hack: use LinkedIn polls to find qualifying users in and beyond your network

I sourced a number of diary participants from LinkedIn polls (that are free). Here’s how it works:

• Create a poll as part of a normal LinkedIn post. Do pay careful attention to the poll wording and options. Consider testing it out with a few people first to get their feedback. Words matter.
• View poll votes to see individual responses. Depending on the engagement from your direct network, a poll can expand to 2nd and 3rd level connections. This amplification is the truly powerful effective of social networks like LinkedIn.
• Reach out directly to individuals who responded with the profile you need. If they’re not first-level connections, reach out in the form of a connection request. This is the key differentiator between LinkedIn polls and other survey tools.

You get multiple signals from LinkedIn polls

Polls are a great method to test out ideas within your LinkedIn network (and those nearby). You get these signals:

• Test general interest in the idea: if you get strong engagement with lots of views, you’re hitting an issue that is topical for your network
• You don’t need a large network personally: it only takes a handful of your network to reshare the poll or vote on it to expand its reach past your network.
• User-level data: when viewing votes, as the people are authenticated and part of your network (and beyond) you get access to the individuals and their submissions.

For me, 148 votes and 10.7k impressions is outstanding engagement from within my network (9k) plus second tier connections.

This was such a great trick for targeted acquisition that I used it a bunch of times over the Knowcast project.

LinkedIn shares details of participants because they’re in my network (or nearby). No other poll or survey tool provides access to individual users without them explicitly opting in.

Understanding who voted what

See the top level breakdown of votes by response. AND critically, access to the individuals who answered that way.

Neither Rob nor Chris were first-level connections prior to the poll but with the poll it cemented an opportunity to make them first-order connections and invite them to participate in the study.

Individual outreach

After the poll closed (a week), I reached out to those who spent 3+ hours listening to podcasts over the past few months. A sample of my connection request outreach is below: note that LinkedIn users may not be frequent check messages that often — you do need to allow one or even two weeks for people to respond.

Thanks to Chris Boud for ok’ing sharing of our chats.

Next I’ll share my experience with the method used for the diary study.