Goal: 1000 Downloads per Day

This is a TWikiAdvocacy question: How can we increase the popularity of TWiki? TWiki is already the most widely used wiki behind corporate firewalls, ahead of Socialtext and Confuence, but unlike the commercial products it is not covered well by the press. In fact, I find the open source TWiki very much underrated by the press. I'd like to fix that.

One good proxy to measure the popularity of an open source product is the number of downloads per day, and its trend. A while ago I created TWikiDownloadStatistics to track the numbers. We currently have around 360 dowloads a day in average. In the last three days after JotSpot's acquisition by Google we had over 500 per day, an indication that publicity brings traffic. My question is, how can we increase the publicity of TWiki to reach 1000 downloads per day?

This is a brainstorming topic, so here we go:

• Create press releases in regular intervals
  • New features, new extensions
• Write case studies of TWiki getting used at the workplace
• Write papers on TWiki and wikis for conferences
• Talk in more conferences
• Write articles for magazines
• Get more well known corporations to write TWikiSuccessStories
• Get one/several corporate sponsor(s) to:
  • Channel press releases and case studies through the pricy PR channels
  • Sponsor work, and have it publicized (it is a fact that the press pays more attention to commercial entities than open source projects)
  • Sponsor talks at conferences
• Build up relationship with reporters of important magazines, news papers, online news sites
• Get more awareness of TWiki in blogosphere:
  • Write blogs that talk about TWiki
  • Get them indexed by blog search engines such as Technorati
  • Write entries on other blogs that (where relevant) mention TWiki
• Form a PublicRelationsTeam that coordinates all this

I am convinced that if we work in this direction we can double and tripple the rate of TWikiInTheNews stories in a short time, and also reach the goal of 1000 downloads a day.

You may wonder why I do not mention better architecture, or more features. I just went to WikiWednesdays two days ago and saw a demo of the new Socialtext 2.0. To their credit, they have a very nice and functional UI, especially with the dashboard that makes it easy to show the "where am I involved?" parts. But then, we are way ahead of them in terms of number of features, new technology used, ability to structure content, and size of community. Well, how come they are much better coverd by the press then we are? In other words, we can increase the publicity of TWiki much more with some focus on PR than on new features. This is not to say, that we should work less on features, just smarter on publicity.

Why does more publicity matter for the development community? More exposure draws more traffic to twiki.org. With this we will get even more contributions for extensions, and also TWiki core. We can recruit more developers. This increases the value of the project. Which is a win/win for all.

Can we reach 1000 downloads a day within the next 3 month? That would be the beginning of February 2006.

Ideas, feedback, initiatives?

-- Contributors: PeterThoeny, RichardDonkin - 03 Nov 2006

Discussion

These are all good ideas - however, they don't really target the blogging community, which is now enormously influential and can really generate significant word of mouth publicity, and is already covering Web 2.0 extensively. So another idea is:

Get more awareness of TWiki in blogosphere

• Write blogs that talk about TWiki
• Get them indexed by blog search engines such as Technorati
• Write entries on other blogs that (where relevant) mention TWiki

Social networking and Web 2.0 are incredibly hot right now, we just need to leverage the huge amount of blog comment and traffic on these topics. This is probably a lot quicker and easier than getting into conventional media, since bloggers already understand the power of collaborative Web 2.0 approaches.

The other thing we should do is to re-position TWiki slightly so that the intro material talks about Web 2.0 - only where relevant, but Web 2.0 is a broad church and I think Wikis are right at the heart of it.

• Good point; I added your bullets on top -- PTh

-- RichardDonkin - 06 Nov 2006

Want to be more popular in the blogger community? What about publishing a TWiki installation package ready for blogging?

-- StephaneLenclud - 06 Nov 2006

Good idea. But then we need to provide broader support than MichaelDaum alone. When packaging, make sure to tell about the user benefits: why would it be better than WordPress?

-- ArthurClemens - 06 Nov 2006

Getting Real was a really nice and easily read book imho - thanks for the link, Arthur. Two chapters I read with special interest were the a powerful promo site-chapter and the keep the posts coming-chapter - in my view they tell us exactly what we could/should do from here.

TWiki is like a lot like showing someone a blank piece of paper "look, you can do anything you want!" - and a lot of people luckily respond well to that.

But of course more could be done for the remaining percentage that deals better with "hey, if you fill out this form the end result you will get is this (only)". I think mainly due to lack of relevant PR persons/ressources our product is often seen as "elitarian" - if you can't figure it out, back luck for you.

Perhaps this is a little off here, but I guess my point is that we could perhaps get more return on our energy by showing off better what we already have got, in preference to implementing still more features? Unfortunately this idea has already been discussed at length in TWikiApplication and others, and so far no one has turned up and signed in on this kind of task - it's a shame but it seems to be the sad but true status at the moment. JotSpot went the office-like way of doing dedicated topic types ("this is a picture page"|"this is a calendar"|"this is .."|etc) - but as we are all, I guess, generalists, narrowing down our scope to this level "just" to fancy the remaining percentage is a tough exercise.

Sometimes it is possible to work together with students on this kind of thing. Perhaps it would be possible to involve some students studying i.e. mass-markets, PR, webdesign, communication or something along those lines into a project along the lines of "Communicating TWiki / transforming collaboration potential into downloads"? Again, someone would have to set their time aside to help in guiding the project / where this kind of setup is possible at all.

-- SteffenPoulsen - 06 Nov 2006

Seems to me like there are quite a few magazines about Linux and open source software out there. Surely TWiki should be able to get good reviews in those pages one way or another. Unfortunately those papers are probably not reaching much of the decision maker audience we need. For the mainstream magazines to write articles about TWiki you probably need to take the journalists for a ride and make sure you don't go for a cheap restaurant. What a diminishing view of the Press I have!!!

-- StephaneLenclud - 06 Nov 2006

see TWiki skin demo using Css Zen Garden for the begining of a series on what can be done with TWiki.

-- SvenDowideit - 13 Nov 2006

Thanks to MichaelDaum for finding this superb image, which sums up for me what is missing between TWiki and 1K downloads per day. While the core functionality of the software can remain stable, it has to be constantly re-inventing itself to attract new users and developers. And that takes huge energy. While many people are still passionate about TWiki, the fire has gone out for many of us as it has become too much of a chore to go through the painful process of idea acceptance i.e. explaining, persuading, coding, demonstrating, burying in soft peat (Codev) and eventual recycling as coal. Instead of being exciting and challenging, it's just challenging. One reason may be the old one that there just aren't enough developers working on the project to keep the idea energy high. Another may be the way that new ideas are usually ignored until their developers give up and go away. As Steffen says above, developers and feature counts isn't the only approach. Some really nice artwork, for example; a revamp of the L&F of t.o., would be a big improvement. Some high quality TWiki applications (I think the Bugs web is still the only one bundled as a contrib, for example). Some "Wow! Look what i can do!" pages that show off existing esoteric features. There's a ton of stuff. The problem is knowing where to start. -- CrawfordCurrie - 22 Nov 2006

Yes, the image is excellent (it's from the "creating passionate users" blog) - I wish we one day will have the luck of having someone thinking like that stumbling into TWiki

Anyway, as they also say in the blog: Life is "trying things to see if they work.", and luckily it looks like some of us are still not entirely finished with figuring out just how TWiki works yet

For my own part I hope that one day I will have some room for doing a new setup around the TWikiVM, my thoughts are along the way of allowing it to be an online showroom, realtime-updated from SVN and "download-what-you-see" (as VM or as source/SVN).

I don't see myself doing a revamp of twiki.org, but having this kind of VM available might work as a "showing-off infrastructure" for someone interested in the task.

I really hate talking about things that are going to be - but look, I fell in again.

-- SteffenPoulsen - 22 Nov 2006

I just received a report from two IT students that had done a series of interviews of users of TWiki in our company. It is in Danish and so far I only have it on paper.

The two major things that they concluded that are interesting for in this context..

• The core user group of 5-6 super users have been a vital factor to the successful introduction
• The roughly 25% of employees who are against TWiki all say the same thing. Using mysterious hard to learn markup that does not look like the final result is a major step back compared to writing in Microsoft office. Ie. it is the WYSIWYG that is missing. Several users were not satisfied with the current 4.0 Wysiwyg implementation as it is buggy and create many problems when used. (25% loves TWiki and 50% never edited a single page in TWiki which is also a bad sign)

And now we are near 4.1 release and noone has done anything or suggested anything related to Wysiwyg.

It is the programmers and very super users that propose things and get them implemented. Not the real users. And it seems that they do not want to let TWiki take them on the ironing board with its current equipment.

-- KennethLavrsen - 22 Nov 2006

Kenneth: In TopThreeFeatures we identified also a fully functional and fast WYSIWYG editor as a key feature for better adoption.

Good discussions on features. I created this topic to talk about how to do better marketing for TWiki, outside of the context of better features. Yesterday I had lunch with a key initiator of the Spread Firefox marketing campain. I will share this later, (now up to building a model rocket with my son, he is anxious to get started...)

-- PeterThoeny - 23 Nov 2006

I have said this before, but the reason why I am pushing TWiki in our organization is because it is a reasonably easy to use web application development platform. We looked at Zope a couple of years ago, and while its user and developer base seemed much stronger, the capability was nowhere near what TWiki had to offer in terms of maturity. We have come even further since then (probably Zope has also, I have not looked).

A wysiwyg editor would be nice, but is not essential for most of our use cases.

A decent inline spell checker is what I get asked for much more often...

-- ThomasWeigert - 23 Nov 2006

My point is that to market something you have to start thinking like a marketer. In this age of saturation advertising, just telling people something exists isn't going to get them queueing at the door. You have to think about the whole front-to-back package; how to hook their interest, how to hold it, and how to exploit it.

Let's assume your interest has been hooked by some viral marketing, and you are a user curious enough to investigate. Look at the front page of t.o. for a moment. It's a bit like reading the instruction manual for a Chinese vacuum cleaner. It's all words, no pictures. Except the blasted TWiki joke robot, which has managed to waddle onto the TWiki.org front page again, despite a long battle to honourably retire it. There isn't even the traditional link to a demo. Interest lost for 90% of visitors. No hold.

The next thing is that in the present age, attention spans are no longer than the time it takes to pick your nose. 90% of people who get as far as the download page will expect a one-click install. Do we deliver?

Assume our user was in need of a complete nasal evacuation, and downloaded, installed and registered without losing the will to live. What next? If you are not a techie, where do you start? There is a mass of documentation - surely I don't have to read that? Where are the tutorials? How do I use this thing? Argh!

The one place TWiki score pretty well is in exploiting interest. An awful lot of people have written and contributed plugins, to the benefit of most other users. But even then, as a community we often seem to go out of our way to demotivate and alienate those people.

Score one out of three, I reckon.

-- CrawfordCurrie - 23 Nov 2006

The difficulty with all these things is that everyone involved has pretty much an idea of what is important for TWiki. How to promote TWiki better, what features TWiki is lacking. But each time we run against barriers: not enough people to do the thing, no champion who will invest his time for one idea.

TWiki developers work in their free time and prefer to work on their own ideas. So it is difficult to target all things we think TWiki needs. But I think we can improve a couple of things. To speed up our development and to decrease friction.

• Decrease the amount of paperwork. If one person is assigned to do a task, other people need to give feedback but should not block ideas. This goes for coding as well as doc work. A too democratic process often leads to mediocre execution.
• If people present an idea, developers need to give quick feedback on how to implement the idea or in what areas difficulties are.
• Give timely feedback. A programmer working for Colas would develop WYSIWYG further in October. But nothing happened. Yet some people knew more about this, so they can help in developing a plan.
• Can we provide donation options for work companies wish to have developed?

-- ArthurClemens - 23 Nov 2006

I think that the point Crawford is trying to make is getting lost. For a start we don't need a better TWiki, as a platform. The best platform (the one with the most and better features) does not always win.

Before or while we are still waiting for someone to pick on the WYSIWYG we need to concentrate efforts on two areas:

• A better TWiki.org, that don't scare away people (what happened with the DocumentationFocusGroup?). This means better documentations, tutorias, how-to's, flash demos, etc. And please, done repeat the old mantra "Go to TWiki web, search the tag". That is just NOT effective, because no one will be there to tell it to the users, plus the users don't want to think.
• A better out-of-the-box experience. Myself, I will consider this goal gone the day I can say to any one in my development team (we're a Java shop, no perlmongers here): "go an install TWiki over that box".

Why this is important? Because if TWiki don't get installed, then there will be no end-users to complain for the lack of a WYSIWYG.

-- RafaelAlvarez - 23 Nov 2006