A quick fix for better engagement on Twitter

Last week I was at a conference and faced with a dilemma: to tweet or not to tweet? The day was full of interesting content, memorable and controversial quotes by the dozen, and some new product coolness showed off on stage. There were large screens throughout the venue showcasing Twitter action that carried the event’s hashtag, and I’m sure many folks who weren’t there in person would have gotten value out of following along via Twitter (not to mention the many journalists in need of exact quotes while they were busily tapping away on their keyboards and working on near real-time stories). So what was the problem? My tweets on any given day tend to cover a wide variety of topics and thanks to this my followers tend to have diverse backgrounds and interests. This conference was smack-dab in the core of adtech — and let’s face it, even advertising professionals tend to get tired of that one pretty quickly. How to engage yet avoid the trap of spewing a stream of content to your followers that they don’t have an interest in?
This was my compromise solution – giving followers a visual cue via hashtag at the start of the tweet

There’s a really easy fix to this with similar functionality already present in Twitter’s product: the much maligned ‘if you start your tweet with someone’s handle that will only be seen by that person and the followers you have in common’*. So we already have the concept of main timeline and direct tweets/replies timeline. Why not have one for hashtags? The principle would be the same: if a tweet starts with a hashtag it would only be visible to people following that hashtag directly. If a user wanted to promote a hashtagged tweet to their main timeline all she’d have to do would be to make sure the hashtag is not at the very beginning of the tweet.

Simple, right?

The main use case is improving the signal to noise ratio on the platform but would also make following topics on Twitter a lot more straightforward, too. Twitter started off as a social network so it would make sense that your friends would be interested in everything you post. As it grows into the world’s platform for real-time communication and engagement it really needs to break beyond the single timeline concept fast.

While Twitter continues to work on attracting new users, it shouldn’t forget features that make the overall experience better for those who’ve been on the platform for a while. We have every indication that the 1% rule will continue to stand so platforms like Twitter really need to make sure they’re catering to the group of users that generates the majority of their content.**

Are you listening, @jack?
* In case you’re wondering why you see tweets that start with a . or some other character, this is why. This remains one of the less-intuitive Twitter features and often makes me wonder if it’s a bug in disguise or a stroke of genius. I prefer to think it’s the latter since it introduces the concept of multiple timelines. Raise your hand if that last sentence made you think of Primer (the movie)  and this visualization.
** This could, of course, be served by a strong developer ecosystem and third party apps that focus on the needs of the frequently tweeting group but we all know how that went down 

Navigating Instagram ads

Screen Shot 2015-10-10 at 4.08.34 PM

Both are ads. But one is not like the other.

Last month Instagram opened its platform to all advertisers and is continuing to develop ad-supported ways of monetizing its global audience of more than 400 million monthly active users. Any platform that commands that size of an audience will automatically be of interest to a wide swath of marketers especially given the granular targeting opportunities that would be available. So how does this all look to the end user today?

The two posts above are both ads – the one on the left is the ‘official’ Instagram ad that you would buy through the Instagram platform. It carries a well-executed ‘sponsored’ logo and tagline in the upper right and for now sports more formats than regular posts do (for instance in this one you can swipe to reveal more photos from the campaign). Is it targeted well? There’s no easy answer because everyone’s Instagram feed is different. For some it’s a social network and a way to interact with people from their social circle. For others, it’s an interest graph meant for discovering and following topics, places, people or things you are into. Then there’s the use-case in the middle where it’s both a mix of social activity and larger discovery, with likely many other sub-flavors emerging.*

The post on the right is the true native ad format on Instagram: it emerged as followings grew and has produced some really interesting collaborations between famous brands and the proverbial ‘woman on the street’ view of the world, coated in authenticity that has been so absent from marketing lately. The brand and the poster arrange a price that’s heavily dependent on the number of followers a poster has as well as typical engagement levels and collaborate on the content of a post. In this case, the poster has clearly identified this as a sponsored post via the #sponsored tag. Most of the comments I spied were complimenting her on her shoe choice (She’s wearing boots and balancing a package of EmergenC on her soles) and were skating right past any mention of EmergenC, the product she is actually pitching.

A couple of things to bear in mind whether you’re a brand, an agency buyer or an Instagram user:

1. Targeting is key

Screen Shot 2015-10-12 at 10.17.57 AM

No thanks

As compelling as the ad monetization opportunity is, from a user perspective poor targeting on a device in your hand, inches from your eyes is so much more jarring than on other media. Platforms like Instagram will need to carefully walk that fine line and make sure they don’t alienate their highly engaged (and growing) user base. If it means seeing more and more ads for app downloads, retargeting, direct response and presumably soon other direct response greatest hits like teeth whitening and belly fat elimination ads (the kind we’re seeing on Twitter) then brands would be wise to withhold their ad budgets for other platforms and channels.

2. What’s an ad and what isn’t

There’s a lot of variance among authors on how to label and identify sponsored posts, ranging from no mention, through ‘I’ve mentioned it on my blog but not so much in the post’ to using certain hashtags like #spon, #ad, or #sponsored. Instagram could help here by enabling a slightly different background or badge for account owners (perhaps similar to their own sponsored one) that would allow authors to subtly distinguish between sponsored and regular posts. Authenticity and transparency go hand in hand.

Another yogi account, same sponsor but a very vague mention of working with the brand.

Another yogi account, same sponsor but a very vague mention of working with the brand.

3.  Find the right spokespeople

If you’re looking to reach an audience with your message someone’s follower count shouldn’t be your only frame of reference. Depending on the vertical you’re in you may be better off with smaller accounts that command more engaged followers. Likewise if you’re an author you want to make sure you’re surfacing partnerships that are not only financially lucrative but also something your followers would enjoy. Companies like MuseFind could help with this type of matchmaking while also providing benchmarks on payment, terms, etc.

There are many underlying challenges of making a successful ad-supported platform where each user can configure their own content experience. There’s no ‘right’ or single way to use Instagram and there will be no right or wrong way to buy and target ads on it either. Here’s hoping that it will remain the domain of brand advertising that supports and enhances overall user experience.




*Mine is the 2nd type: it’s an interest feed and I mostly follow a group of yogis, athletes, high fashion brands, photographers, street artists & art aggregator accounts. In that sense, the creative looks good although somewhat puzzlingly it’s showing me men’s clothing first (the last two images in the carousel out of four are female).

**MuseFind were part of the most recent cohort at Entrepreneurs Roundtable Accelerator and I had the pleasure of mentoring them when they were first accepted. I’m a big fan of their leadership team and their platform is wonderfully straightforward and easy to use. Check them out.



A glimpse into the future

Screen Shot 2015-09-02 at 3.19.43 PM

There’s a very interesting discussion on AVC this week that fits squarely into a few topics I’ve been thinking about rather intensively this summer. To that end, I’m borrowing Fred’s title for this post. The background reading is this piece by Bloomberg’s Justin Fox and the money quote:

“In general, developing countries are making the switch from goods to services much earlier in their development than the U.S. and Europe did.”

This is a tectonic shift. We are used to measuring success and status via material goods accumulated but the switch to valuing services across more levels of the socio-economic spectrum aims to disrupt that. What if people in the developing world can skip the car ownership step altogether in favor of, say, on-demand drivereless cars, the way they’re skipping telephone land lines and desktops in favor of smartphones? How will that change their overall consumption patterns?

While the crystal ball is a great tool to have whenever considering the future, there are 3 large change agents to how we do things today that are visible and already underway:

  1. How we innovate: today innovation is globally largely still synonymous with Silicon Valley. I think (and am betting on) the defining characteristic of the next 50 years of innovation will be the emergence of diverse local hubs. The initial start will likely be mimicking successful companies (like what Rocket Internet is doing at scale) but within a decade we should start seeing local successes especially in the larger markets first (e.g. Turkey w/ 75MM people and mobile penetration rates in the high %60s).
  2. How we learn: with many localized variations, we generally subscribe to the concept of elementary, vocation, and higher education. It’s interesting to observe how much variation there is between countries on what should be covered in elementary school.* With the availability of excellent learning materials for free or negligible fees via MOOCs perhaps that basic curriculum can be toned down to a more general set of skills (logic, reading comprehension, foreign languages, mathematics, statistics, and other principles of analysis) augmented with very deep dives into more specific topics of interest. Tech aptitude is rapidly becoming the new literacy and the sooner our formal education systems are able to reflect that the better off our kids will be.
  3. How we make things: advancements in 3D printing are really intriguing to monitor. While we may still be a few years out from true market adoption, for the first time in recent memory we can contemplate what producing goods locally will look like. There’s an entire new economy hidden here: for example, how do you monetize and incentivize product designers**, what does the cost structure for more complex products look like, are there premium printer options/materials to consider, etc.

Most of our assumptions about the developing world are based on how the now-developed world has grown. Technology gives the developed world the opportunity to skip a few steps on that ladder and as a result come up with stronger, localized markets. We’re looking at a very interesting decade that will potentially warrant a tome of itself in the history books*** of the future.


*In Serbia, by the end of elementary education we were expected to command basics of calculus (math was mandatory), advanced physics (4 years of study), chemistry (3 years of study), 1 or 2 years of Latin among a variety of other subjects that my American colleagues didn’t cover until university or in some cases very advanced AP classes in high school. It made for an interesting transition from one educational system to another and back. 

**Most today seem to fall flat in the hobbyist category which is great given that most people interested in computers in recent memory were considered hobbyists until the first dot com boom/bust and the beginning of consumerization of computers. 

***Assuming we still have the concept of books in the future. And that we haven’t accidentally blown said future up.

Through the eyes of a viewer – a simple test of TV’s addressability promise & the perceived value of video ads today

Which button should I press?

Which one is the easy button?

Television has always been a predominantly passive experience: get comfy on your couch, tune in at the right time (primetime!) and a bevy of good content is about to pour your way especially if you don’t touch that dial. Fast forward to today where remote control dials have largely been replaced by a more active consumer experience: instead of wait and casually browse, we search and operate on our own timetables, on demand. Catering to this shift in viewer behavior is the main advantage of various VOD and OTT systems who are more closely tailored to the search vs. browse than old school MVPD interfaces and remote-control-accessed menus can ever be. But advertising budgets still command a heavy premium on the linear TV side as opposed to its digital cousins, although that gap is shrinking daily. While both are technically addressable media, the flexibility of ad targeting and buying on the digital side is far greater and the promise of granular audience targeting is there for the taking. But never mind the industry view; what does an average viewer think about all this?

To attempt an answer, I decided to concoct an experiment to compare the experience of watching the same content on traditional TV and via online VOD. The premise was rather basic: impersonating your average viewer, I’d watch an episode of the Daily Show with Jon Stewart in the evening when it first airs on cable and then again over VOD the following morning, comparing the ads and my overall viewing experience. For VOD, I opted for Hulu over Comedy Central’s own streaming portal since that seems to be more analogous to how a MVPD operates (e.g. Hulu aggregates content from multiple networks, vs Comedy Central’s service only streams its own; this struck me as a fairer comparison). To make things more equal I watched both of these on the same screen: a beautiful 65” Samsung SmartTV (in the one case via STB, and in the other Hulu Plus over Apple TV), thus putting to bed the old adage that VOD is for small and secondary screens only. So on a Thursday night I settled in on the couch comfortably and started fishing around through hundreds of channels to find Comedy Central.

The results surprised me: on the linear side a total of 25 different ads came hurling my way and I couldn’t discern any particular targeting pattern. Movie promos and trailers led the way with 6 – all for upcoming releases of different genres. The most confusing category was auto: the 5 ads included high-end marques like Jaguar and Infiniti, middle-ground Hondas and Toyotas, and a ‘sponsored by Kia’ spot that kicked off the episode (and was immediately followed by the Infiniti ad). Judging by the ad load Daily Show viewers must be in market for all cars all the time. The rest were fairly run-of-the-mill national campaigns for telco, consumer electronics, insurance services, rounded out by (only?) 1 local ad and 1 PSA.

Screen Shot 2015-07-07 at 5.25.19 PM

Hulu’s interface gives viewers information on how many ads they have left and option to select an ad they’d prefer

On Hulu the following morning (as soon as the content became available), locating the episode took 4 quick taps*. Instead of 25 there were 13 ads, split into a pre-roll and then 3 blocks of 4 ads per. Each block was clearly labeled showing how many ads remaining, countdown to when content will resume, and a menu option to choose a different ad if I wanted. It feels like the viewer has a choice – I’m not passively being fed a stream of ads I don’t particularly care for; I can, at least in theory, do something about what I’m watching. It’s also interesting to note a very different line-up of what I was served: movies and autos were notably absent in favor of food, beverage, services, and pharma. Again, I could not discern a clear targeting pattern other than the ads were clearly (and correctly) skewing female in both creative and topic (e.g. pharma ads were for feminine health issues which is interesting and arguably quite valuable since the audience of the Daily Show tends to skew towards the young male demographic).**

At a glance: breakdown of ads shown against same episode on STB vs digital VOD

Screen Shot 2015-07-07 at 5.14.19 PM

So what conclusions can we draw? From a pure consumer perspective I’m finding it really hard to argue in favor of the linear TV experience. From an industry perspective it doesn’t readily seem we’re taking advantage of the promise of addressability. After all, when it’s the same content and same screen but such a different advertising experience, why not decide what to watch based on the ads alone? It’ll be interesting to see how MVPD’s own attempts at OTT shape up (like Dish’s Sling service) and if they can prove to be truly competitive with the OTT natives like Netflix. One thing is clear: viewers have more video entertainment options than ever before***, and the bar is not measured on content quality alone, but ease of discovery, access, and overall viewer experience (ads included).


*One to switch on Apple TV and 3 to find and play content.

** Here’s a good rundown of Daily Show average audience size and changing demo, all through the lens of Jon Stewart’s impending departure

*** The South-Korean performer Psy of “Gangnam Style” fame at the time of writing commands 8,061,763 subscribers on YouTube. The 2 year old “Gangnam Style” video has been viewed a total of 2,370,147,492 times (that’s more than 2 billion views with a capital B).

Mary Meeker’s most interesting slide

There are 197 slides in this year’s Internet Trends presentation by Mary Meeker and her team — or the gospel of the Internet, as I lovingly call it. If you’re reading this you’ve probably spent the better part of yesterday afternoon dissecting the deck. As usual it’s filled with very interesting insights, some well-deserved recognition of market-makers come market-leaders and all manner of statistics that will quickly become part of funding pitches, strategy outlines, go-to-market decks, and assorted assets. But there’s one slide this year among many great ones* that truly stands out.

Slide 153. This one:

Screen Shot 2015-05-28 at 8.19.06 PM


  • It’s about removing friction from mundane real-life tasks. This is what ‘Uber for ______’ should ideally look like.
  • It ensures higher quality of input (and output) data. No more paper forms and indecipherable handwriting**. Data input becomes structured, with error-checking at each step. Pre-populating fields can ease form completion and submission.
  • It’s a natural extension of a chat platform. It doesn’t re-invent the wheel or require you to download a myriad of ill-designed separate apps. It’s already relying on a system that most everyone knows how to use and has on their phone. By continuing to add functionality its utility continues to increase (especially if there’s a way to seamlessly pre-populate and repurpose repetitive information like address, basic personal information, etc).
  • Its logical next step will be more automation. “Your license is expiring in 1 month. Would you like to renew it?’ can be an alert. With the quality of on-board phone cameras, there’s fewer and fewer reasons why one wouldn’t be able to snap a passport or visa photo on the phone. With a few taps and a swipe you could skip the queue and have your documents dispatched to you. Now wouldn’t that be lovely? Then let’s all wait for the day when all of those documents actually live inside our mobile devices.
  • Its cost-saving and efficiency increase could be staggering. Bureaucracy is expensive and no one likes it; not even the bureaucrats, it seems. While rudimentary data input until recently typically required a digitization step, with the penetration of mobile devices*** so many common requests can be created and processed digitally right off the bat.

This is not a definitive list but the mere presence of this slide and the existence of this screen within the WeChat platform is an incredibly exciting step towards a more seamless future. The one in which mobile technology truly becomes an enabling utility for the currently connected and unconnected worlds alike.****


* Full deck, which you probably already have, can be found here (along with previous years’ versions all the way back to ancient history, aka 2001):  KPCB Internet Trends

** I had to fill out a form with a pen and send it via fax yesterday, May 27th, 2015. There’s no earthly reason why anyone should still be using a fax for anything. Very few things make me as grumpy as faxes do.

*** That’s on slide #5. We’re globally at 73% population penetration and rising. For comparison, the most recent UN study on global penetration of improved sanitation (aka % of people with, at a minimum, access to a 20th century version of a toilet) is at 64% (but the data for this was sourced in 2011 so hopefully we’ve improved dramatically since then). 

****This futurist can dream. Also, in this scenario, artificial intelligence doesn’t kill us all. 


‘Hey, wouldn’t it be sweet if we got some of that tasty, tasty earned media?’ said The Brand. ‘I know’ – The Brand’s social media manager and evil sidekick jumped into action. ‘There’s this show called House of Cards that everyone’s talking about, and its 3rd season is going live today – let’s say something and piggyback on their hashtags for maximum win. Our product has nothing to do with the show and we’re not funny or anything, but who cares, let’s do it!’

…. Aaaand scene:

Screen Shot 2015-02-27 at 9.04.31 PM


Hey brands — yeah, you. The reason people like social channels is because of authenticity (yeah, we could debate this till spring but let’s not) — when you try stunts like this you’re really just shooting yourself in the foot.

Go Commando!