Development

The month was spent trading emails with Alex.  Our understanding of the application and how to use phpBB has evolved although, as it has always been, it was a painstaking process.  I was slow to realise and accept that the application could be developed as a series of phpBB mods.  Our greatest achievement, therefore, has been to settle the technical approach:

1. maintain a media module (media interface)
   • CMS (i.e. phpBB) can be updated (e.g version > 3.0.2) relatively easily (key benefit is security)
   • CMS can be changed relatively easily (e.g. to vBulletin)
   • isolates (most) application logic from the CMS
2. design custom css/html template
   • enables more flexible (simpler) styles
   • may include customized index page
   • may include customized mod pages
3. create mods for each feature (or set of features), as needed:
   • projects are well defined (i.e. module or version)
   • projects are contained

Theory

Watched an interesting TED lecture which, I think, relates to this application.

Development

Spent $500 plus $200 bonus on development.  Work done by Ukrainian programmer called Alex (handle is ‘Vadini’).  Used Scriptlance.  He asked to see complete specification before quoting.  It felt like high risk but, it turns out, it is low risk because the scope for alternative interpretation is always great.

Before Alex, I nearly committed $1000 to a probable fraud called Greentech12.  I Googled them and discovered this warning: http://glowhost.com/forums/general-chit-chat/greentech12-1108.html.

I described the application on Scriptlance in this way: ‘The application (front end) brings people together in a specific way in order to communicate via the bulletin board (back end)’.  We are using phpBB but my intention is to only modify its styling.

Name

Invested time searching for a better name.  Received advice from friend who came up with ‘Ice Patrol’ for her forthcoming documentary on Antarctic research vessel.  She slammed my candidates; none were compelling or communicable: “People must know how to spell it on hearing it and have no trouble remembering it.”

Look2Look still popular with family and friends who have been following progress from start.  I let LooktoLook back onto the market earlier in the year and, after rekindled interest in Look2Look, discovered LooktoLook had been acquired by a Swiss person offering to sell it for about £1000.  If Look2Look ever earns money then I will buy it back on revenue, not capital (i.e. money earned by the application, not me).  I acknowledge that writing about it here will effect the price.

Mootka is still relatively unpopular.  I encountered ‘Ka’, ancient Egyptian for ‘spirit’, in the novel that I am reading.  It felt prescient.

I used the ‘Friday Quiz’ at work to gain feedback from my unsuspecting colleagues.  Look2Look came out on top but I do not like it.

Style

At the end of the month my template looks like this:

It is a fragment of the design in order to protect ideas.

Theory

This diagram illustrates where search-on-search (new) may fit into existing user search habits (old):

A quote from Wikipedia that has a bearing on this project:

Others suggest that the term social software is best used not to refer to a single type of software, but rather to the use of two or more modes of computer-mediated communication that result in “community formation”.[6] In this view, people form online communities by combining one-to-one (e.g., email and instant messaging), one-to-many (Web pages and blogs), and many-to-many (wikis) communication modes.

Source: http://en.wikipedia.org

A quote from Wikipedia describing Google Groups.  My solution is very similar:

“Google has introduced a new concept in Google Groups to reduce redundant threads and therefore helping moderation of forums. Basically, the idea comes from the fact that many users do not bother to search a forum and directly create new threads to seek an answer to a question. When the user types a new thread subject it brings up similar threads automatically on the side. This helps keep the number of redundant threads (or the overall forum pollution) to a lower level as users that neglect to search for a topic and are posting a thread may find the answer to their question as they are creating the new thread. In essence it’s a search combined with creating a thread. Instead of having to search the forum, and then creating a thread if there is no satisfying answer, this duplicate thread prevention allows the user to go and seek the answer to a topic. In Google Groups [this feature uses an] AJAX type interface and only draws from the subject field of the thread to find similar threads.”

Source: http://en.wikipedia.org/wiki/Comparison_of_Internet_forum_software

Design

I had a few design revelations:

• Groups application can do the job of the dating application.
• I do not need to list all user searches.  Google does not list all websites.
• I can make landing page like Google search’s landing page.

Copy

Have long struggled to find a name for ’saved searches’ with comments attached.  Have tried ’searches’, ‘groups’ and ‘moots’.  New contender is ’search-groups’. 

Labour

Matt delivered…

…then disappeared without explanation.  Reckon he has had enough.

With Matt’s absence, I have turned to Scriptlance.

Means all my ideas go public but I get much more control over development.

Competitors

New competitor called Chanced It spotted by Marjorie.

Closest technique to Look2look so far.  My bet is that they will face the same issues as Look2look.  Could be wrong; their design, styling and copy is way better than anything I managed.

Press

Read some great comments on The Register by Anonymous Coward.  

http://www.theregister.co.uk/2008/07/31/it_jobs_good_or_bad/comments/

Here is his take on hiring IT staff:

The good thing I’ve noticed is companies have realised they can vet cvs quicker than agents, so they have one woman in personnel stick out a job advert and catch a thousand cvs. At this point some 2 techies short list over their lunch hour removing all those who, can’t spell, think they were using ODBC when they were using OLE DB etc, have hobbies such as “Socialising”, “Music”, etc — eg coming in with a hangover or tired, and texting mates all day. Then they’re cut down again, removing all the premenopausal women, but keeping all the women you’ve got with children all over 7, one of these women will know exactly how to solve the company’s woes because she’s been there 20 years, and is brilliant but noone listens to her because she’s female and doesn’t have any ambition. Then you throw 50% of the rest in the bin - who wants to employ someone who is unlucky?

At the end you’re left with a few dozen, so you get them in, and reject all the tall project managers - tall people are universally overestimated so they’re not going to be as good for the same level of promotion as someone who has got promoted despite being a dwarf. On the techie side you look at their career history for success. The brilliant never fail inspite of the management, so they’ll have worked on a load of difficult, but still successful projects.

Weed out those who want to use every latest bit of technology, they leave before they’ve finished, and hey presto you’ve got your men.

Theory

Some thoughts:

• This application brings people together based upon common expressions of interest.
• This application reduces the cost (money, time or status) of joining or forming a group.
• Gaming is important because it binds people without spilling blood.
• Voters can be more idealistic than their politicians because they have less to lose.

Use Cases

Some examples of how the groups application can be used.

CASE 1: “beer London”

I fancy beer on Wednesday night somewhere in town.
I write ‘beer London’.
I want sometime this week (not any week).
I only want to know if any of my friends have thought of it too and so I restrict searches to all my contacts.

CASE 2: ”Prince 2 next month London”

I must study for a really boring professional exam.
I was going to study with my brother but he’s had to go abroad for work.
I write “Prince 2 next month London”.
I find a bunch of people in the London area thinking the same thing.
We chat away, exchanging notes on the best course supplier.
After a while, I get to like a few of them and invite to become contacts.

CASE 3: ”footy ealing common”

I see guys playing football on Ealing common on my way back from work.
They look pretty good at it.
I wonder if anybody in the area is up for some less formal ‘five aside’.
I write “footy ealing common”.
I find several groups.
A couple look quite serious but there’s a smaller group of 2 people who, judging from their comments, seem more like me.
I join the group.
I delete my original search because I want new entrants to join this new group and not my old one.

Strategy

Went on sunny holiday and had a think. Main conclusion I came to was that there is a popular application still to be made out of ‘search-on-search’.

I have reconsidered how the ‘groups application’ works and have asked Matt to start building it.

Look2Look is still online but is, effectively, shelved. I have asked Matt to fix the bugs and put it on Facebook but it is lower priority than building the ‘groups application’.

Design

New design guide:

1. Look2Look was for…everybody. Everybody actually means nobody. This new application will be for my friends. If they like it, others will too.
2. Look2Look needed everybody to know about it before it could work (i.e. big critical mass). The groups application will just need a few users (e.g. my friends) for it to work.
3. There is no point investing in advertising, fancy styling, logos or adaption to Facebook or mobile until the application is being used and talked about by my friends.
4. Any effort to ‘look good’ must not be at the expense of accessibility or usability. People care more about whether it works on their old computer, slow connection, tiny screen or Braille display than ‘prettiness’.
5. The more platforms the application works on, without compromising content or function, the more useful it is. All else being equal, the traveler will pick the application that is as effective on his Blackberry as it is on his laptop.

Labour

I attempted learning MySQL and PHP but lost heart when, after 7 days of trying, I failed to fix one simple bug in Look2Look.

Matt is willing to build the groups application but his time and interest are quite limited. I intend to improve my management of Matt by:

• After the initial brief, let him get on with it without interference.
• Instead of change requests, provide feedback from the user’s perspective.
• Although the design can wander, the goal remains to create something my friends actually use.
• Provide well considered feedback by email, not chat or voice.
• Establish and rigorously fulfill a weekly checkpoint meeting (currently Monday morning UK or NZ time).

Name

Currently, the name for the ‘groups application’ is Mootka. It’s the best I can do for the moment.

Thoughts about names in general:

• Names are important. Facebook paid $200,000 to change from theFacebook to Facebook.
• It’s the most important thing about you application that gets passed from person to person.
• You can change your name without upsetting your business because it is easy to forward people from your old URL to the new URL. This means you can go live with a name you’re unhappy with and improve it once you’ve acquired a better one.
• It’s nice to have .com but not essential. Most people find websites via Google rather than typing in the URL.
• Don’t change a name unless a new name comes along that ‘blows it out of the water’.

Here are some rules that I have tried to abide by when finding of a good name:

1. The domain name is available (.com)
2. Is a made up word (we want to give it our own meaning) or does exactly ‘what it says on the tin’
3. Doesn’t mean something bad in another language or slang (e.g. Hungarian for ‘fuck off’)
4. Short (two syllables or less)
5. Memorable
6. Likeable (people not embarrassed to pass the name on; worth noting that I had trouble remembering Google when I first heard it)
7. Trips off the tongue (French have trouble with ‘th’)
8. Phonetic (no silent letters - ghost/gost, plough/plow)
9. Is a verb (’to google’ versus ‘to amazon(?)’)
10. Incorruptible (’halfkin’ corrupts to ‘foreskin’)
11. Unique (will quickly rise to top of search results)
12. No numbers acting as words because it sounds cheap (e.g. Furniture4less, Phones4U)

Mootka

“It’s shit” (Jason)

“It’s brilliant” (Matt, Andrew, Nick)

“Sounds like a new thing that come on the internet thatyou might want to look into” (Shanks)

Pull a face (Mum, Alex, Joe)

“Sounds like Russian vodka” (Ashworth)

“Still like ‘mootka’ more than other options. I am too used to feel this word ugly now, but it still sounds extraordinary. Like the term ‘reciprocal search’ too.” (Tomoko)

“It’s OK” (Ashan)

“Sounds more interesting than others” (Cathy)

The primary purpose of the groups application is to bring groups of people together based upon common expressions of interest (i.e. searches).

It’s different from Look2Look in that, for any one search (string of keywords), it brings more than two people together.

It is different from other group applications insofar as it doesn’t really distinguish between creating a ’search’ and creating a ‘group’. This should reduce the barrier to creating a group.

Press

Oliver Lindberg interviews Jim Buckmaster, CE of Craigslist, in .Net magazine, August edition 2008.

.Net comments:

Online classified advertising site Craigslist is run more as a public service than for profit.

Maximising revenue has never been part of the company’s motivation and it’s how it survived the dotcom bust.

Craiglist only charges for recruitment ads in the San Francisco Bay area and 10 US cities, and brokered apartment rentals in NY City.

Craiglist doesn’t even have a logo, but instead uses the peace sign as a favicon.

Jim Buckmaster comments:

“It’s important for us to be a profitable concern because we don’t want to borrow money. We don’t want to have to sell shares.”

“Everything we do and every decision is made from the perspective of the users of the site.”

“We don’t consider [text ads or banner ads] because users aren’t asking for [them].”

“We’re not fans of unnecessary complexity.”

“Our design criteria really only includes things lke usability and accessibility.”

“Any time you add a newer user interface technology, you tend to exclude some small percentage of people from using your site [...]. Another design criterion is page load performance.”

“Design, from a user perspective, does little more than distract and obstruct them from what they’re trying to do.”

“There are thousands or tens of thousands of businesses that offer classifieds, maybe hundreds of thousands. Looking at what other companies are doing is mostly time wasted. Our time is best spent concentrating on what our users are asking for and trying to do the best we can for them.”

Craigslist facts:

• in business for 13 years
• 10 billion page views with 400 million using the site each month
• worth $5bn
• 567 cities, 50 countries, 6 languages
• 25 staff

Review

Released first version of Look2Look in March 2006.  It flopped.  Put it down to poor appearance, mainly.

Spent from April 2006 to June 2008 trying to make the application more compelling by experimenting with the way it looks, its pitch and the way it works.  Not achieved.

Some consolation from seeing 6 (ish) other websites attempt the same thing over the same period but, apparently, without success.

The conclusion I must draw is that this particular use of the search-on-search technique (i.e. two people looking for each other) is a lame duck however you cut it.

Development

Final design for this month looked like this:

Right at the start of the month, Nick got in touch. I showed him what I’d done…

He said, ”I personally think the layout is far too simplistic, I suppose its important to balance between usability and unique design, although I do think you are on the right lines. Need a hand with the design at all? ”

This blue one was greeted by Mark Osgood, a colleague, with a flat and emphatic, “it’s worse”.

Competitors

Spotted a new competitor.

Development

Had very helpful input from colleagues, mainly Chris Hocking.  Stripped out lot of features as a result.  He also recommended a friend of is who is a graphics designer called namkwancho.co.uk.

Matt’s disappeared for a while due to illness.

This month’s experiments (having trouble letting it go):

Strategy

The application has gone back to the Look2Look URL.  Thought is that Look2Look may not work but it’s done so let’s put it out there and crack on with groups application.

Development

Throughout development I have struggled to explain what my application does.  I have wanted a pithy one liner.  I think this might be it:

“Use keywords to find each other.”

It requires some follow on explanation but, having tried this line in casual conversation, I find people get the gist straight away.

Here is the latest design attempt.  It feels ‘elegant’ and colleagues who have seen it like it.

Good things:

• Lots of white-space.
• Centred on the screen but flexible width.
• Feels like a serious ‘tool’ - not a gimmick (inspiration from Google and Craigslist).
• Minimum use of colour by employing an ‘eyefeed’.   Putting primary (mainly) colours in a prominent place (top) on the page seems to satisfy the eye and thus lets me get away with extremely simple styling through the rest of the page.  I think Google’s logo enables Google to employ the same trick.
• Using a lot of browser defaults.
• Only uses one column and mostly uses browser defaults, therefore, will render well on broader range of browsers and load to wireless devices quicker.
• Colleagues liked it.

Bad things:

• The title visited hyperlink clashes with the eyefeed.
• The logo at the bottom clashes with the eyefeed
• The list of searches look bad without a strong left margin.
• The links do not all operate in the same way (Chris Hocking’s observation).

Competitors

Phil Shanks (colleague) alerted me to new website doing very much the same thing as Look2Look called findmyship.com.  Phil spotted it in Metro newspaper.

Emailed the owner and, much to my surprise, got a reply.  Owner was Brendan based in Dublin.  Good bloke.  He said the site had been up for a year already but he has only just started to market it.  Here are some of the things he said:

On what he is doing right now:

“Yup this is my baby, although there are more bodies getting involved in particular around the marketing side, and we’re about to do a tech upgrade, providing I can justify the spend. (To date it’s just been a ’spare time’ project really).”

On his previous experience:

“Really though it’s an experimental ‘toe in the water’ thing. I’ve been a keen follower of the social networking trend for years. We tried launching an online dating site around 8 years ago back here, but were beaten to the punch and pulled back.”

I am not the only one who thought there was a market:

“Since then I’ve just watched all these social sites grow and felt there was a market for a product like ours.”

Where he got the idea:

“For me it was a girl I met in a taxi rank at 3am on a Saturday morning. We got on famously, really clicked, but she was so suddenly whisked away by her friends who pulled up in another cab. We had that, “ooh what now” moment but everything went so quick that no contact info was shared and the ship sailed on!! I went back and said to myself, ‘ok, how do I find this person again?’ but obviously couldn’t. “

How he is marketing it:

“I’ll keep you posted on our progress, but really everything about what we’re doing now is in around the marketing area and specifically around viral marketing campaigns. The Metro link was a result of an email blitz on a range of radio station and magazine emails based in Ireland and the UK. I wasn’t expecting anything to come out of it but it popped up on the radar in a couple of places which is all we want really. We’ve also been doing some local late night flyer and card hand out advertising which has generated some interest too.”

My kick off was Shepherd’s Bush.  His is Dublin:

“I agree with you completely on the local issue too. Our initial concentration is on Dublin, and specifically key venue locations within Dublin. The biggest threat is the ‘run out of marketing steam/budget’ issue too I think. We need that user base to really kick off any type of advertising revenue in the longer term.”

The problem of ‘critical mass’:

“In talking with friends and family I realised that this was so much more common than I originally thought, and the concept of findmyship was born. I mean in theory, like your site, it’s a great idea, but depends ENTIRELY on the brand being a common, household name! If everyone knew that such a site existed and what its primary purpose was then I’d have no doubt such a site would be a raging success. Educate the masses!”

On motivation:

“At the moment it’s all just a labour of love, albeit a strange warped and perverted love!”

It is remarkably rewarding to have met someone who is under apparently similar circumstances, has had similar experiences, drawn similar conclusions and made similar decisions.  Good luck Brendan.

Labour

I attended phpLondon (http://www.phplondon.org/wiki/Main_Page) event on advice of Alex Tingle (whom I went to meet at his home in March). I demonstrated the product to two programmers I met there.  They thought it was cool but were not in a position to help with development.  Met programmer who was local, apparently capable and charged £30 per hour.  But I am sticking with Matt for the moment.  Hoping I can rekindle his enthusiasm by moving onto the groups application.

Nick got back in touch offering to work for free in exchange for future profit share. Nice offer and nice to know that someone else still has faith in this whole enterprise.

Strategy

The site has been up for three months. I have tried to generate interest with online advertising. None forthcoming.

I am beginning to sense that it is time to let go of this incarnation of search-on-search (two people looking for, exclusively, each other) and to have another crack at the ‘groups’ application.

Why now? Firstly, after the above changes to the design, I believe the current application is as good as I can get it without further user input. Secondly, I think that I have figured out the groups application. Last time we tried figuring out the groups application, we got stuck. I have done some more thinking and I may have cracked the problem. No details yet.

But, first, I want to try a few more things (below) and I my quest for the perfect pitch and design hasn’t quite dissipated yet.

Development

Latest design attempt:

My latest, perhaps last, change requests:

1. Public discussion

The idea was to attach a discussion thread to each search. Up to this point, a discussion thread was only created when one user contacted another. This thread exists independent of the search and is ‘private’ because only the two participants can see it. I saw that online classified advertising sites like Gumtree had public discussion threads linked to each search (i.e. ad) and thought a similar arrangement for Look2Look would make the application more compelling.

2. Personal messages

Each classified listing on Gumtree consists of one line of hyperlinked text. If the user clicks on the hyperlink then the full ad opens. I noted, with the addition of a public discussion thread, that searches on Look2Look were starting to look like Gumtree classifieds. I also noted that Gumtree ads enabled the ad owner to write a personal message or expound upon the basic listing. I thought this feature on Look2Look might help make the application more compelling. Matt and I had difference of opinion about execution but resolved by end of month.

3. Contacts

We got rid of a module that listed the latest chats by, instead, ordering the contacts in the Contacts module by activity (most recent at the top of the list).

4. Logged Out Searches

We have struggled how to persuade the new user to sign up to the application. Previously, our method was to ‘talk them into it’. Again, taking my cue from Gumtree, I thought a better way would be to make the application immediately interesting by enabling new users to see and search for searches without having to sign or log in first. Initially, our concern was that this would undermine the point of the application - which was to record what people searched for - because people could search and leave no trace. However, I felt that there would still be a strong incentive to log in if the search facility on the logged outside was much less powerful than the facility on the inside. In other words, we had ‘quick search’ on the outside and the full cross-matching and filtering form on the inside - which saved your searches.

Strategy

I am still figuring out how to get users. I still believe the key lies in how it looks and works.

What is the look that I am after?  I discussed this with a colleague (Lee) who has graphic design background. We concluded that, because the application was somewhere between a search engine and an online social network, it should give the impression of an ‘online social network that feels like a search engine’.  Like me, he thinks that achieving a design that communicates this will make or break site.

Development

Development slow.  Matt does not see where the project is going.

Below are some screenshots of my design attempts for this month.  This one attempts to mimic Google’s page layout:

Discovered that dropping title cases and some punctuation greatly improved the application’s appearance.

Matt came up with the idea that whomever initiated a chat could provide their name to the other person. It’s a good idea and I should have let him run with it.

Instead of screenshots showing the user what the application looks like once logged in, we are debating allowing the user to search when logged-out.

Promotion

Advertising online in order to encourage someone to sign up.  No luck so far.

Realised that I can have ‘midwife pages’. A user coming from, say Gumtree, lands on a ‘midwife page’ that pitches the application, specifically, to Gumtree users.  It could only be accessed from a link on the Gumtree site.

Intellectual Property

The Subject Matter Application was due to expire.  Had to decided whether to pursue the patent process or drop it altogether.  Short history of advice:

Dec 2006 - bunch of PWC executives said ‘patent your idea’.

Mar 2007 - submit preliminary application which establishes priority.

Aug 2007 - bright technical bloke says ‘there’s nothing new in your idea’.

Feb 2008 - bright strategy consultant says ‘without the prospect of real revenue, even if the patent was valid, its worthless’.

I wrote to Keltie (my patent agents) asking them to confirm the costs of going ahead with the patent process.  Here are some excerpts from their reply:

“Drafting costs of approximately GBP 9,000 (excluding VAT).”

“PCT filing costs of approximately GBP 3,000 (excluding VAT).”

“Filing of a US application from the PCT will be largely the same as a direct filing and will be approximately GBP 3,000 to GBP 4,000 including our service charge and our US associates fees, together with the official fees.”

“Unfortunately, it is not usually the case that patents go straight through to grant, and so it is very unlikely that no further costs would be involved.”

“Typically, each round of examination will cost approximately GBP 500 (for a very simple response) to GBP 4,000 (for a very complicated response incurring our time-based charges and a local US agent’s charges for reviewing and filing the response).”

“Typically, costs of GBP 1500 to GBP 2000 are not unusual for responding to a US Office Action, and one or two rounds of examination are not uncommon.”

“If you have not disclosed your concept, you may feel that it would be preferential to withdraw and re-file the UK application at minimal cost. However, if you have disclosed your concept out of confidence this option is not advisable.”

Matt thought £20K (rough total) was ‘insane’.  Piers (strategy consultant) pointed out that the USA awards patents to ‘first-to-invent’ which means my original Subject Matter Application, with it’s early submission date, still counts as useful evidence in the USA.  Michelle (Keltie) felt there had been insufficient public disclosure to void the patent and so I decided to re-file for approximately £400 - which gives me another year’s grace.

Research

As a research exercise, I studied Google Groups. It does pretty much everything I had hoped my groups application would do which leaves me wondering whether there is anything that I can add.

Development

Someone pointed out that ‘all applications look shit without users or content’.  Didn’t stop me trying…

Research

I opened a Twitter account out of curiosity yesterday.  Good name, good design but a success because it is both instantly rewarding (like any other blog) and novel (currently, nobody else offers the ability to publish by text). 

Good article about the founder and its foundation from a December edition of the Economist magazine:

Competitors

I was looking at Gumtree’s ‘sponsored links’ with a view to advertising there and discovered a new competitor.

Look2Look

This is the year that I launched my first search-on-search application. It was called Look2Look. It failed.

People came to the website but did not login. I concluded that this was due to bad pitch and/or design and I worked to improve both.

Meanwhile, new competitors kept appearing which led me to believe I was onto a good thing but also made me very keen to ‘crack the nut’. I threw lots of time, energy and money at it.

Groups application

Concurrently, I spent a lot of time thinking about how the second or ‘groups application’ of search-on-search will work. Had a go at making a prototype but realized, on starting the build, that I had not thought it through enough.

Strategy

The product and pitch are still not good enough yet.  Despite this we are back online for the second time.

Development

Was using 2 column layout to accommodate inline help but was unhappy with it because I do not like using floats (reduces browser and small screen compatibility).

Various attempts to make it look more appealing. I want to use browser defaults as much as possible but this limits what I can do. Taking inspiration from Craigslist, Google and Wikipedia. 

Maybe I need to show the site is ‘alive’. Could do this by showing something topical -  for example, winter photos in winter time or usage statistics.

Research

Read some interesting stuff about web design in Mind (Scientific American).  Web-designers have 10 seconds for their homepage to snare user,  sentences need to be very short - no paragraphs and, if you want people to pay more attention, make something harder to read.

Strategy

Main effort is still design.  Wish I could program.  

Development

Efforts this month met with conflicting responses: “uninspiring” and “simple is good”.

Search-on-search

What problem does search-on-search solve?  

It is worth searching for things that do not or hardly exist yet.  Examples are ‘beer next Tuesday’ or ‘anyone else thought about playing football on the green on Tuesday’s’.

Very specific searches are worthwhile.  With traditional search, ’salsa dancing Kensington’ is a worthwhile search but, with search-on-search, ’salsa dancing Kensington last Tuesday guy in red shirt at the bar’ is a worthwhile search.

My interest is in social applications but there may be other problems search-on-search may solve.  I know, from speaking to an ex-Detective Inspector, that the Metropolitan Police have an application that uses something like search-on-search on their intranet.  Helps them solves crimes.

What types of search-on-search are there?

I think there are 3 types.  The simplest matches one field to another.  The next cross matches my first field with your second and your first with my second.  The third is a combination of the two so for finding each other, you cross-match my ‘I look like’ field with their ‘you look like’ field, their ‘I look like’ field with your ‘you look like’ field and then you both reference the context, ‘Waterloo Thursday’.

Marketing

New Scientist magazine (2 or 3 issues ago) described how the weight and extent of a political candidate’s personal influence on the people around him accurately predicts how many votes he will win overall.

This chimed with an interview with the founder of Craigslist that I once read. He said that he answered every email that was sent to him.  A personal email from ‘Craig’ of Craigslist’ is more likely to turn you into a loyal user than a newsletter.

My conclusion is that, if you can find the time, answering email personally is a great way to create a user base.  It’s my guess that, once Craig had got to ‘critical mass’, he could ease up on the emails. 

Craigslist’s old site.  Pretty simple for the 5th most visited website in the world.  Fancy design does not seem key after all.

Craigslist’s upgrade.  Not much different.  

Social theory

Found a paper by a guy I saw in the Royal Society called Martin A. Nowak, a mathematical biologist.  He talked about indirect reciprocity.  Here is a key line from the paper:

“In a fluid population, in which most of the interactions are anonymous and people have no possibility of monitoring the social score of others, indirect reciprocity has no chance. In a socially viscous population, in which people know each other’s reputation, cooperation by indirect reciprocity can thrive.”

In other words, indirect reciprocity cannot thrive in modern society and, therefore, neither do the feelings and behaviors that it induces.

Competitors

Also, new site called Icebrkr.com (see scan below).  Same problems and solutions I had on launch. Thanks to Patrick for spotting the article in FT Money.  Here is how they describe their technique:

1. Arrive at a bar for a drink with your friends.
2. See someone you fancy?
3. Click on the SMS we sent you for instant access to our mobile Internet site.
4. Browse photos of other members in the very same bar as you.
5. Found the person you fancy?
6. Click their photo to send them a friendly message.
7. If they reply ‘yes’ then go over for a chat.

Development

This month, tried adding photos.  Will that do the trick?

Or maybe bold primary colours is the answer.

Search-on-search

Invented an alternative name to search-on-search: ‘reciprocal search’.  Is a reference to ‘indirect reciprocity’ or ‘reciprocal altruism’.  

Google search (below) suggests the term exists but not much literature.

Research

Another article about the shallowness of online social network relationships. 

Development

My best design effort was reviewed at the end of the month by a trusted friend.

She thought it looked ‘rubbish’.

Removed chunk of unnecessary functionality - more accurately, merged register of chats into contacts.

Strategy

My friend’s advice, an accomplished business woman, was to not release now.

Copy

Indecision over application’s name. Look2Look sounds like “dating” or “daytime TV chat show”. Mootka is short but forgettable and ugly.

Facebook

Less concerned about ending up as a Facebook application after I heard some guy did very well by it.

Online social networks, like Facebook, provide status reassurance, status calibration and safe dating. Are they the last in the line of social applications? No, if better methods found.

Article: The battle for Facebook by Stephen Foley, 26 September 2007, The Independent.  Useful line: “Dollars follow eyeballs”.  Sums up my business plan.

Included in the article is a quote:

“Facebook’s decision in May to open its platform to web developers (who can build unique mini-applications, or widgets, which users can populate their Facebook profiles with) demonstrated some of the same attributes which have helped Google become such a popular platform among both developers and users,” said Mr Jayant. “As a result of this open approach, many companies are now investing in building customized Facebook applications, further fueling an already fast-growing ecosystem.”

Open is good.  Closed is bad.

Press

Strategy

I have run out of money.  Dropped Nick (design), kept Matt (programming).  

Labour

Matt is getting busier with other customers.

Copy

Have bought ‘Mootka.com’, a less ‘dating’ type name.

Development

Now I am doing the day job and learning how to do the design work myself.  This involves reminding myself about HTML, learning CSS and getting to know an open source graphics application called GIMP.

Facebook

Facebook has changed the state of play. 3 or 4 months ago, it began letting anybody join (not just students). Last month (ish) it invited external developers to build their own add-ons.

I met a Facebook founder (Matt Cohler) last year and his opinion of my application was that it was a ‘feature’ of an application, not an application of itself.

Becoming a Facebook add-on feels like a bit of a come down.  

Instead, I am considering expanding what that application does (search-on-search engine) and eyeing up a less competitive platform (social network for, specifically, mobile).  

2005

I was in a cafe. There was a girl. We pretended not to notice each other. English reserve, cowardice or the exigencies of daily life resulted in failure to communicate. The opportunity was lost and so was she (London is a big city).

I began wondering how to recover the situation. The idea of search-on-search popped into my head. Search-on-search is a search technique: you search for other people’s searches. With no experience and no programming skills, I decided to set about building a web application and, eventually, a business based on this idea.

August 2005 to February 2007

Built two concept demos using MS Access and ASP. Paid colleague at work to do this. Mocked up user interfaces using MS PowerPoint. Finding the right programmer (Matt) took from August 2005 to November 2006. Got in front of a panel of PWC consultants and their advice was to patent the idea. Learnt about Intellectual Property law and agents (the good and the bad). Found start-up capital in family (£10k). Declined capital from couple of friends (£5k each). Talks with potential partners (UniVillage) did not come to anything. Spent lots of time figuring out what could be done with this new concept. Prepared many presentation slides to show to investors. Got interview with one high-net-worth person. Told to come back when I had something to show him.

Had to choose between developing an application that helped two people find each other by cross-referencing their descriptions or an application that used search-on-search to form groups. Chose the first because it was my first idea, it seemed more compelling at the time and I had not quite figured out how the groups application worked.

March 2007

In 2005, I had an idea. It might be a good one.

This blog records what I am doing about it.

I keep a record in order to account for my time (people sometimes ask and I sometimes reflect).

I keep it on the Internet in order to save me accounting for my time verbally and to force me to explain clearly (I cannot get away with the sort of scribbles that you will find in my notebooks).