We were thrilled to get a chance to talk to the folks at MarketMuse's Content Strategy Collective a few weeks back in an hour-long "Ask Me Anything" session. Our CEO Kingston Duffie and I chatted with content professionals at length about a variety of pressing topics surrounding engagement, publishing and more.
We thought that some of the topics discussed would be relevant to our publishers as well, so we decided to share it here - lightly edited for clarity. Enjoy!
Annabelle Hopper, MarketMuse
Buckle your seat belts....this #ama is getting started! Welcome, Kingston Duffie and John Greely from SlickStream! We are so excited to have you all here today!
John Greely, Slickstream VP of Marketing
Hi everybody, we're so excited to be here!
Kingston Duffie, Slickstream CEO
Hi! Glad to be here.
I'll kick things off with my question! Thanks for being here!
Slickstream is my fourth Silicon Valley startup. I’ve been lucky. I’ve seen a lot of winners and losers in general in the startup world.
I have to admit that we’re still figuring out who the winners are going to be. We see Elastic and Algolia as search providers who are succeeding at the top of the market. But we don’t see much competition where we are.
We’re excited to see what MarketMuse is doing — bringing new ideas like AI into this market.
hi kingston and john! thanks for doing this ama. i'm wondering in what ways can our team consistently be increasing user engagement? what has worked for you in the past, and what are some areas to explore?
I think one of the things we've heard from a lot of our customers in the blogging space is the importance of resurfacing older but still valuable content
In the sense that posts, studies etc. have hours of hard work poured into them, only to have a limited shelf life until the next thing comes along
So even aside from the work we do to resurface it when it's relevant, one thing our publishers are constantly doing on their own is finding ways to update, adapt, tweak and repurpose that content for the moment, and they can get significantly more value from it
A couple of lessons we learned… 1) publishers think they know what viewers want, but aren’t always right — better to give viewers way to easily find exactly what they are looking for; and 2) less is usually more.
A good example: a publisher is often inclined to push their most popular content. But it turns out that offering recommendations that have high affinity to what the viewer has already been reading is much more likely to produce clickthrough.
I mean Kingston and I could talk for the full hour about this one! I do want to touch on another component here though, which is just prioritizing user experience first
We can all get caught up in our KPIs, for email signups, downloads, bounce rate, or whatever your current focus is. And it can be easy to optimize for those specific goals
But I think people often forget that every bit of affordance you give to a specific goal comes with a tradeoff, which is that overall user experience. And in our testing to date, we've noticed that increased engagement doesn't just come from pageviews/session, it comes from sessions/user as well. And fostering those return users is a tricky business! (edited)
HI Kingston and John - Thanks again for doing this! Do you see any big content trends for the second half of 2020 and heading into 2021 that you think we should pay close attention to?
Interesting question. Unlike MarketMuse, we focus our attention on how to keep and engage viewers that our customers have already succeeded in getting to their site. Obviously, better content is always what works. But we’re not really in the business of guiding publishers to topical stuff. Having said that, one thing we do know is that publishers who specialize are much more likely to have been engagement.
In other words, if you “shotgun” your content, you may get SEO, but you won’t get
It’s amazing to watch people navigate around a site when there’s lots of really strongly related content.
I think it's hard to predict trends for such a volatile year, but we actually did some research recently into past content patterns in a lot of our sites, particularly in how they related to revenue: https://www.slickstream.com/blog/covid-food-blog-revenue-impact
That was fascinating for us, since it confirmed a lot of anecdotal reports of massively increased traffic with really abnormal patterns
hi Kingston and John -- Thanks for joining us. (product/packaging brain) Your pricing is so simple. How confident are you in that approach? Ever feel that you leave money on the table? or, are you happy with the simplicity? How do you approach this with larger clients that may need more service or special requirements?
Oh gosh, pricing!
The simplicity is absolutely a feature rather than a bug, and it helps us operate as a fairly lean team. We don't have a sales staff, so the need to negotiate/upsell/haggle to stay afloat would have a pretty sizeable impact on us
I'd say there have absolutely been instances where we've left money on the table, and many more where our minimum pricing was too high for smaller trialists. But we designed that model with two core components in mind: transparency and fairness
We have a big advantage because we are able to prove ROIs that are fairly constant over wide ranges of traffic. So pricing to traffic has been an easy fit for us.
Everyone pays the same rate, that rate scales with pageviews, as do our costs. And I think that policy has been pretty readily defensible and accepted by our users
But Kingston can definitely point to many instances we've had larger prospects make product requests that soon turned into full-fledged features for us!
We share in our customer’s success — both instantly (in immediate pageview increases) and long-term (in steadily increasing return visits)
Yeah. I’m a huge believer in letting customers direct our product development.
But that only works well when you prime the pump with some innovative ideas for them to react and respond to.
For example, are any of your familiar with Web Vitals? https://web.dev/vitals/
Google has been pushing this “standard” for measuring site performance. We have the benefit of being on every pageview and so are collecting all of this data. Now we’re discussing with publishers how to help them actually use it.
Steven van Vessum
Hi Kingston and John, thanks for taking the time to do the AMA! Question that’s relevant now that Google has officially announced to make Web Core Vitals ranking factors: how big is the impact of implementation Slickstream on the site, and do you have anything special planned to further minimize its impact on speed?
We're using several of the pieces of an offline solution (CacheAPI, IndexedDB) plus just-in-time design. For the person arriving the first time, we get out of the way and lazy load. But on any subsequent fetch in the same session or a later one, we have everything we need stored inside that browser itself, so that we load along with the page itself -- without depending on any roundtrips to our server.
We did a big release several weeks ago where we replaced our embed code on customer sites with a new one that made a radical improvement on performance. This used some advanced browser tech to make it possible to load MUCH faster.
This was essential for us because as a SaaS, we are an overlay on the site — loading in parallel with the site’s content.
Thanks for being here! What are some of the main challenges you have faced in terms of creating a better user experience? How did you adapt and solve these challenges?
First of all, we have a really GREAT guy on our dev team when it comes to UX. That’s the first thing — good people.
But we also follow a few core principles in all of our design. First one — we need to seamlessly integrate with the publisher’s look-and-feel. And SIMPLE, SIMPLE, SIMPLE.
We try things and are willing to drop them if they’re not right.
Yeah just to jump on that look and feel answer too
One of our biggest emphasis points is onboarding, and we spend a fair bit of time getting that right. If we can make Slickstream look like a part of a user's site rather than an add-on, it makes all the difference
It's also important to note how much communicating we do with those users. We have multiple touchpoints throughout our trial period (and after) where we're actively soliciting feedback
So when the support team or myself hears from a user about a UX issue, something they like, or something they want, that feedback goes straight to engineering and we discuss it as a team
We definitely take advantage of still being a small team. A new feature sometimes arrives on live sites within days and sometimes even hours of requests.
Heya! Question when you have time.
How do you see the future of communication changing in the coming year?
Well, “communication” is a big one. Depends on how narrowly you define it. If we’re talking about web tech, I’d say that the trends are definitely towards increased interactivity. Rich client-side is steadily coming.
If you mean “communication” in a bigger sense, I’d say that Covid is changing everything. I think we’re seeing a lot more real-time and near-real-time communication.
what does that adhoc user issue discussion process look like? how do you determine the value from user requests vs. roadmapped efforts?
We depend a lot on the collective instincts of our team. After a few minutes of discussion, it’s usually easy to tell whether a new idea is a “no brainer” to implement, or is filled with uncertainty.
Also, we work hard to maintain a balance between short-term and long-term dev investments. We put about 25% of effort in continuous re-engineering. Another 25% in quick-response features. And the rest in broad efforts.
For example, we’ve got a huge waiting list for our new Stories product. https://www.slickstream.com/stories But we’re taking it slow because we want to make sure we get this one right before we are all-in.
Hey Kingston & John ! What are some website eye-sores for you? An engagement trend that doesn’t add much value or, on the flip side, obvious tactics that haven't been adopted yet?
IMHO we see very aggressive popups trying to get people to sign up and/or accept browser notifications. I think these can easily turn off viewers. Remember that we think long-term engagement is much more important than short-term. If you annoy your viewer as they’re leaving, they are less likely to come back.
That’s just one example of a counter-productive engagement strategy, I think.
We try to be agnostic to the 'best' tactics, but I think there are a lot of sites who would do well to think about affordance. That is, what specific thing do you want your visitors to be drawn to first. If a site has a full-screen opt-in popup, then a large floating video ad, it's unlikely you're going to be creating a positive user experience that incentivizes a return visit
We’re delighted to see that the customers we’ve been with for more than a year are seeing really good year-over-year growth. We’d like to think that we’re part of that success.
What types of data do you think are the most important to be collecting? What are your tips/tools/ways of measuring big data? How do you go about interpreting, analyzing, and implementing changes based on data? Thanks! :sparkles:
We watch every pageview — even every scroll and click on every pageview. And we collect a HUGE amount of data. Many terabytes. It’s amazing how valuable it has been to have all of that data. It’s hard work to maintain, but definitely worth it. We use mongodb and we have lots of tooling we use to analyze, query, and aggregate it.
I could probably talk for an hour on this subject along. But let me think about most important tips…
1. Keep everything for as long as you can. You don’t know what you’ll need until you want to query it.
2. A good data consolidation strategy is essential — balancing short-term versus long-term info needs.
3. Live in the data. Wallow in it. You’ll be amazed what you find if you just keep querying it in different ways.
4. Don’t overwhelm the customer. We have a LOT more data — even in summarized form — than we show customers. We tend to wait for a customer to ask for something — perhaps many times — before we decide to add it into the UI.
Well, I’ll shut up. (I love data!)
As a tangent from it while Kingston talks big data, for content publishers themselves though, it can all be pretty overwhelming
I'm going to thread it here. It's so difficult to parse the signal from the noise sometimes, and I think the most difficult thing when evaluating content success is getting discouraged by not seeing large immediate uptick
Overall engagement isn't about viral wins for the most part, it's about subtle compounding gains over time
If you can make it just 1% more likely that a one-off visitor becomes a return visitor, then you've just slightly increased your traffic for every future post too.
Those aren't just platitudes or cliches either, it really goes back to that sessions/user stat I mentioned earlier
Hey John I’m curious to know what types of elements do you utilize in creating user experiences?
What is it that makes the user experience unique? How do you do that?
To elevate those engagement rates.
That's a full team effort of course, since the product ideas need to be implemented right
But I'd say of the tools we offer for engagement, they each serve a different purpose that some sites value more than others
The one that drives the most pageviews is the filmstrip, which isn't dissimilar to any content slider you've seen before. The quality of recommendations is critical there, and it can generally drive 1-2% pageview growth
The favoriting functionality is more about social proof, and focused thus on bounce rate and retention
And the search tool gets people to the content they're specifically looking for
We didn't invent any of the individual concepts, nor did we discover the value of them, but I think it's important to understand that engagement isn't just one thing to one type of user. It's an overall integrated experience on your site that makes sure every session is successful
Let me jump in to say that we invest heavily in client-side interactivity. If you look at our widgets, everything is instantaneous. As you type, you’re seeing ALL of the search results change in real time. As you look at a bunch of recommendations, you can infinitely scroll them in place.
We've got three minutes left, and I want to expand on a point from my thread with Stephen Jeske for the group at large for posterity
We designed Slickstream to tackle 'engagement', which is generally the flipside to the acquisition coin
Getting people to your site matters, and what they do on the site also matters
But engagement isn't just one thing. A direct user vs. a social user vs. a search user might want very different things, and the key is delivering them a successful session
So a search user might reach your site, see 20,000 favorites on the page, and say 'oh this will work' instead of bouncing
A direct user might vaguely remember the post they were looking for and search for it again, but the search tool needs to be able to find it
And a viral social post visitor might not care at all about your site, but if they see an interesting second article in the filmstrip they'll click it too
It's not one thing, it's all the things. And I think a focus on what makes a session 'successful' for each type of your visitors is the most important thing to think about for content marketers in general
Obviously John Greely thinks a lot about this stuff ;)
And...that's a wrap! HUGE shout out to Kingston Duffie and John Greely for a spectacular #ama! Awesome insights and detailed feedback! Thank you all for participating, hope to see you again next time