A Brolik Periscope Experiment: The Agency MVP

What is Periscope?

Periscope is a relatively new streaming video app (owned by Twitter) that lets anyone broadcast live from their mobile phone.

Why is it important?

Periscope may very well change the world. Imagine first hand video accounts of protests and riots, streaming video from the front row at the Super Bowl and more. Opening up “journalism” to the masses, for better or worse, will change everything.

What does this mean for a digital agency?

At Brolik, we like to stay up to date on all new technology, especially if there is a potential social or marketing component.

In the case of Periscope, the social and marketing potential is really what interests us. We decided to do a small spin on the traditional webinar by offering a bit of lecture and some Q&A live on Periscope.

People are still figuring out the medium and what viewers want to see, so here’s our take.
Watch the video below of our first live stream from a few weeks ago about minimum viable product in an agency setting. Then, I’ll expand on some of the great questions we received.

What we covered for the experiment

When we started planning the Periscope broadcast, we needed a topic. We talk and write a lot about the minimum viable product (MVP) approach to building software, and we deal with a unique side of MVP as an agency using the approach for client work.

We’ve learned some valuable lessons about using the MVP approach as an agency, and we knew there were a lot of questions around the subject. It was the perfect topic for our first Brolik Q&A on Periscope.

In case you aren’t inclined to watch that whole video, here’s the information we covered and some of the questions we received.

Why is MVP hard as an agency?

The key here is education. It’s hard to educate clients and manage their expectations. The process itself is different, new and sometimes scary, so it makes client education very important.

It’s not only important to educate on the mechanics of the MVP process. It’s also important to explain that it will be scary and may seem counter-intuitive at times. The more your client is ready, the smoother the agency MVP will go.

Read more about building client projects with an MVP approach in another article I wrote.

Why do clients like MVP but not trust it?

The short answer to this is that almost everything about the MVP process is appealing but completely new to a client. It’s like nothing they’re used to, so they don’t trust that it will work once they dive in.

There are five specific things that help alleviate that distrust:

1. The product needs to always feel polished and have no dead ends or broken functionality.
2. The product needs to do its “one main thing” exceptionally well. That’s the minimum viable part.
3. Things should be manual until they need to be automated, but everything should feel automated (even when it’s not).
4. Implementing change needs to be easy so that you can show your client how fast you can pivot when necessary. (This will alleviate a lot of stress and allow them to more freely “test” ideas.)
5. All stakeholders need to be educated early. The hardest ones to get to are probably the CEOs and business owners, but they may be the most important people to have on board.

What’s the best way to collect user feedback?

The easiest way to find out what users are doing is analytics and other automatic metrics.

From straightforward analytics like Google Analytics and Visistat (mentioned as a top three sales prospecting tool in this article) to more advanced data analysis like our friends at RJ Metrics provide, data can show you a lot about what your users are doing.

Beyond that, there are user surveys like Survey Monkey or Google Surveys.

Of course, one of the best methods of getting user feedback is speaking with actual users. It’s hard to beat this method, especially for the really great feedback you can get by “reading between the lines” of what people are saying.

If you’ve got a budget, try some testing with real users. The Nielson Norman Group explains a few methods here.

What’s an example of a manual process that seems automatic?

The most obvious example, and the one I used on Periscope, is the example of Magic.

Magic users text requests for “anything that’s legal,” and the app magically takes care of their request.

While it appears that a computer or robot is handling your request in a very advanced and automatic way, it’s actually fielded by a human who’s using the exact services you’d have to if you didn’t use Magic.

Magic is just a huge MVP. If the idea works, they’ll automate things over time. For now, they can learn what to automate before spending their resources automating blindly.

Are smaller companies a better fit for MVP?

Size isn’t necessarily important for the success of an MVP project, but the approach gets more challenging as the companies involved get larger.

I’d argue that MVP is the best approach for any project, but ultimately, there are advantages and disadvantages to MVP in general, as well as specific advantages and disadvantages based on company size.

As I mentioned, company size factors in because instead of two or three people needing to buy in and understand the MVP process, 20 or 30 people have to.

Besides the education component, when there are more stakeholders (or more affected team members), there are always going to be more considerations, more feature requests and more opinions in general.

All of those things are MVP killers, so the more there are, the more challenging it is to stick to the minimum viable features. The trick is to frame every question, request and decision against the main project goals or the goal of the current phase.

Is it dangerous to say MVP is cheaper?

MVP is only cheaper to get something to market. It’s really important to remember that in the end, MVP isn’t a cheaper way to finish a project. It’s actually just a cheaper way to fail (and then pivot, of course).

If you ran the same project using a traditional model and an MVP model, the difference would be in the final product, not the final cost.

The MVP product will get to market sooner (for less money initially) and be more in line with what customers want and need.

Are there any pitfalls to MVP?

As you may have picked up on… of course! Most of the pitfalls revolve around education and expectation, as I’ve talked about throughout this whole piece.

The results

In the end, we saw some encouraging results.

We had 136 live viewers, which was more than we predicted. Since it was our first experiment, we only sent a few tweets ahead of time. We barely promoted the broadcast at all.

This leads us to believe that if we promote a broadcast, and if we broadcast regularly, we’ll see a lot more viewers.

We also wondered how many questions we’d get. The realtime response was enough to keep a talk going for more than twenty minutes, so that’s a win in our book as well.

All in all, our initial experiment was a success, and we’re encouraged to set up more serious benchmarks and put effort into Periscope in the future.

On top of all that, it was fun!

° ° °

If you have specific questions or concerns about your MVP project or if there’s anything you’re nervous about, feel free to contact me on Twitter (@drewbrolik) or email me (thomasa@brolik.com).

Here’s the original Brolik Blog article that we based our Periscope content on: How to Build an App and Minimum Viable Product (MVP).

We plan to Periscope more. If you’re interested in watching the next one, download Periscope now, and follow us on Twitter (@brolik). We’ll announce the event a few days early, and we’ll announce it again right before we go live.