Blog

graph.me Relaunch Coverage

graph.me Relaunch (as Facebook Timeline)

This is how our relaunch could look like on a Facebook Timeline, some bits and pieces skipped. Will there ever be a timeline for Facebook pages? We think it’d be cool, and we’re not the only ones.

Right now our Facebook page looks like any other page: facebook.com/graph.me. Like us if you like us!

Relaunch

The new graph.me

graph.me has become a powerful platform for quick polls, votings and fast feedback with plenty of insights. At its heart it’s still graph.me, but we have learned and reworked so many things that we want to introduce you to a brandnew product.

The branding has changed and goes better with our new positioning. And we redesigned the whole user interface, following more common design patterns. It’s new under the bonnet, too: Flash seemed more and more like a dead end so we decided to relaunch graph.me on a completely new (HTML) basis.

All the changes and new features in detail:

Demographics Being able to filter poll results by personal or demographic characteristics is what makes graph.me really exciting. It adds a lot more value to the results when you’re able to select women between 30 and 40 from the voters, or your very target audience. We have improved the filters: You can choose from a greater variety of predefined characteristics and you can now also define your own.

Individual duration Now a poll can run 20 minutes, 7 days, 3 months or forever. You set the individual duration when creating a poll, can edit these settings later or simply hit start and stop.

Widgets We put a lot of emphasis on our most popular feature. Introduced last winter, graph.me widgets enable you to simultaneously place a poll in as many sites as you want – on your blog, your website, in your shop or on a Facebook tab. Until now you had to newly embed every single poll you started on graph.me, even if the widget had the same size, position and design. Now there’s a second type of widget: Every time you start a new poll, this widget is automatically updated and your most recent poll shows up, with the option to include a full poll history. Voters have to provide demographic data just once, it‘s then being used for other polls in the same widget and session. We improved the looks, too. You can adjust more design parameters and save custom designs. And a poll widget can be spread really easy now, as the embed code is available right in the widget.

Poll types? Multiple choice. You can still choose from different poll types besides multiple choice questions. Use a 1-5 scale for likelihood, rating or sympathy and get the trend displayed in a special gauge. Or ask people what they’re willing to spend on your product and get numerical results displayed with statistic benchmarks and a distribution chart. We have improved the results view on graph.me and in widgets.

Results undercover Say you want to use graph.me for getting feedback from your customers. If you don‘t want anybody to see the results right away, simply activate the option for not disclosing the results to poll participants. If you‘re happy with the outcome and want to make it public later on, simply change that setting.

Reports, charts and raw data Download a PDF report for your polls, anytime. It summarizes all the results and comes with segmentation based on demographic data plus a time-wise distribution chart of answers. Furthermore you can have all charts from the report as image files – to use these in your next presentation, for example. And for the number crunchers: The raw data is available as an Excel file.

Editing polls Once a poll was started you couldn‘t correct or add things to it. The only way was to delete and restart your poll. This has changed: Now you can make adjustments to all parameters, add, delete or correct answers later anytime.

Inviting participants With switching to HTML we have integrated more options for sharing a poll. Post it on Facebook, Twitter, Google+ and Xing or send an invitation with just one click to ask your friends, followers and circles to participate. And graph.me loads much quicker now when accessed via an outside link.

Comments new graph.me comes with Disqus built in. So anybody can add his or her personal free text opinion after voting.

Freemium. Good news: There’s still the Basic Account that gives you the most important features for free. The Pro Account is coming with all the new things and is aimed at those of you with a bigger audience (these features also being available for just one poll – pay as you go).

The brandnew graph.me, summed up:

It looks better, comes with a bunch of new features, introduces a freemium model and is based on a solid new architecture for all the world changing things and products we still have on our mind.

Rework

Lessons learned

What were the lessons learned after successfully launching graph.me?

1. Building up a closed community was tough, especially with lacking social features and incentives. Requiring people to register was too big a hurdle for many, even more so when just voting on a poll. So that was one of the first modifications we made to graph.me, though it disturbed our initial idea of aggregating data a bit.
2. Polls and trends are not totally unlike, but usecases turned out to be. These two elements being under one roof have never been easy to communicate after all and too much at a time.
3. Users liked graph.me‘s quick, easy and nice looking polls, especially with demographic filters. Tracking trends not quite, our approach was too broad to make it really useful and we would have needed to rework the whole thing. Creating a separate service would have made more sense after all. But we focused on what was going well: polls.
4. We were critized a lot for building a 100% Flash app. Though we had our reasons and by the time things were not like they are today, we have to admit it was a wrong decision. Unfortunately none you could straighten out on the quick so we had to wait for new graph.me.
5. With getting away from a closed concept we received frequent feedback whether it’s possible to embed our quick polls on other websites. This feedback came from bloggers and site owners who wanted to question their readers or customers right were they are – an obvious target group we were going to focus more with relaunching graph.me. As a first step we introduced graph.me Widgets last winter.
6. We were pretty sure of the value of aggregating data and deriving findings from graph.me content and had a couple more revenue channels on our mind, but our business model was mostly dependent on building up a huge userbase first before graph.me could be monetarized. That was not exactly what investors wanted to hear, being rather risk-averse in the days of financial crisis and looking for ventures with revenues from day 1.

Rework.

We needed to switch to HTML sooner or later and wanted to give graph.me a new approach, so a relaunch made sense. So we sat down, discussed, planned and started all over again.

Our objective was to refine our concept of quick, easy votings and flexible, meaningful feedback. Based on a more solid and down to earth business model. The result is a (r)evolutionized product which is targeted more at professionals than the first version, but still has Social Benchmarking in its genes.

We were glad to have valuable feedback from first pilot customers like Fidor Bank AG. Also blogs like netzwertig.com had used graph.me til then, thanks to Martin Weigert for objections and suggestions here, too.

We believe, starting and establishing a new business means monitoring, testing, learning and constantly reinventing yourself. Especially when you haven‘t done it before. You need some perseverance and a strong team that believes in what you‘re doing and that is willing to walk the way together.

And this is where we‘ve come so far: The new graph.me.

Rewind

How Social Benchmarking became graph.me

First there was an idea: Social Benchmarking was supposed be a new online platform for people sharing personal opinions via quick polls and tracking numbers from every day life via trends. To find out where you stand compared to a certain peergroup. By aggregating all this data we wanted to find out things like “Men from 20-30 who go running three times a week in Munich, prefer BMW to Mercedes and spend 2k a year on their holidays“.

Having nothing but an early concept paper we successfully applied for a German government grant called EXIST Gründerstipendium. As one of the first web startups to get this one year support from the Federal Ministry of Economics and Technology (BMWi) and with additional help from the LMU Entrepreneurship Center we could start turning our idea into something more tangible. We were greenhorns, so the learning curve was steep. Also from a technical angle, as we decided to build all of graph.me in Flash. Then there were a myriad of features we packed into our specification, thus the whole thing took a bit longer. Sure, we‘d do it differently today.

If you didn‘t have an invitation to our closed beta all you could see was our Bavarian inspired waiting queue (yes, we are based in Munich). Meanwhile many were excited as we pitched at TechCrunch Munich and were awarded best startup – with public launch still a while ahead.

Then the time was right for officially launching graph.me so we got up on a TechCrunch stage again, this time at GeeknRolla 2010 in London. That our presentation was the one the audience liked most made us proud and gave us the right mood, considering that some well chosen startups from all over Europe came there for pitching and launching.

And there it was, our first product, live for the first time. Since then we have monitored, modified and learnt a lot. And started to think how we could do it better – with an all new graph.me.

Read about our lessons learned in the next part.