When Intercom had just one product, we held one live product walkthrough per week, to anything from 1 to 50 potential new customers. Even if that meant getting up at 6am in Dublin to demo to customers in the Southern Hemisphere, we never missed a Wednesday demo. The benefits were clear – customers who attended a demo were less likely to churn.
But product demonstrations were also a resource sink. Even with one product and a small user base, we still required someone to update the website, send reminder emails, host the demos, field questions, and report on attendee numbers. This overhead grew dramatically as we began to tailor our demos to each new product, and our customer base grew.
Presenting live also came with technical challenges. Your product won’t always look or behave the way you want it to. Your mic won’t connect. The tool you’re using to screenshare may not operate correctly. And you’re at the mercy of internet connection issues; even a brief delay in loading a page can knock the wind out of your sails mid-demo.
It was clear we wouldn’t be able to scale personal demos for every product. One approach we experimented with was screensharing a slide deck of annotated product screenshots. It was simple for any of the team to create engaging slides, which could be quickly edited as our product grew. It wasn’t the most polished process, but moving from a live product walkthrough to slide deck meant our demos were richer, snappier and easier to deliver.
But we were still required to deliver the same live demo every time. We needed something even easier, and something repeatable so other team members could deliver it. We needed an approach that required less overhead, without impacting the quality of the demo.
As Intercom has grown, we’ve produced all sorts of video content; help videos, best practice guides, and promotional material for new product releases. This back catalogue of video inspired us to rethink how we demoed our product. What if we could pre-record part of the demo, but have a live Q&A and discussion too? So we decided to try out an approach we called hybrid demos.
We would chat with attendees at the beginning of the session, briefly again halfway through the video demo, and then once more at the end for the live Q&A and discussion.
Hybrid demos still felt “live”; attendees knew an Intercom expert was still there to answer their questions. And thanks to tight editing and some basic animations, these demos were faster and richer too. It meant we could deliver consistent quality to attendees every time.
It also afforded us more time to run a solid Q&A session. We rarely like keeping attendees more than an hour, so our old process meant we often had to cut sessions short without getting to all their questions. Now that our pre-recorded demo gave us a concise 20-30 minutes of quality content, we had 30-40 minutes to do a live Q&A and discussion. A lot of customers come to webinars solely for the Q&A, so they felt far more involved and satisfied that their concerns had been addressed.
With our hybrid demos, we saw technical overhead drop significantly, while demo quality, and time for Q&A all improved.
We also sent recordings to all attendees so people could quickly watch the content again or share with their teams.
Clearly, the introduction of video worked. Our demos were standardised, dependable and effective, and we carried on with this model for over a year, analysing their impact.
We knew that trial users who watched our demos were 11% more likely to retain past the two week trial period and become paying customers. And they were subsequently over 6x more likely to activate than customers who didn’t watch them. We knew our demos were making a real difference to our bottom line. But, try as we might to get as many people as possible to sign up, we still only saw about 3.5% of new customers attending a demo. We needed this content in front of more eyeballs, and our reach was maxed out with our current model.
As an experiment with one of our product demos, we decided to try directing customers to watch a product demo, entirely on-demand.
The advantages for the customer were clear – they could watch whenever was convenient for them, which made a big difference for customers across different timezones. They could also easily share the material with their teams. And rather than dedicating an hour to watch the demo, the core content was now only 10-15 minutes long.
The major disadvantage from our perspective was the lack of a live Q&A session, and the absence of a sense of human connection you get from a demo. That meant there was a risk of lower engagement, even with more viewers.
To mitigate those concerns, our on-demand videos encouraged viewers to use our messenger for questions, and pointed to other help content on the site.
The results were amazing. So much so we went ahead and took this approach with the rest of our demos. We soon had 7 times more people watching our videos than had attended our demos. But the real question was how would people engage with our on-demand content without a real live human connection, and what would they do afterwards?
Somewhat to our surprise, we saw a lift here too. Users who watched a demo on-demand were now 14% more likely to retain into month 2 than those who had not watched a demo. So that’s a 3% overall increase from the old world of scheduled live demos.
But far from supplanting live demos and Q&As, this on-demand approach complements the other demo models. It has allowed us to clarify the specific areas in which live events and video excel. We continue to use live webinars for when we’re looking to encourage customers to take specific actions, where a live setting means we can tailor the content and Q&A to suit the specific needs of both us and our users.
Meanwhile, our on-demand demo videos are getting shorter and more focused too. We want them to be agile, and exist in lots of different contexts. For instance, our Sales and Customer Support teams can create video playlists they know will benefit their clients, without any of the content which may not be relevant to them.