Trends: Story1 – How do you solve the mapping problem effectively?

I want to do a few quick posts to cover a range of activities in certain areas – trends, if you will, where some of the trends are old and some are newer.

Other budding Entrepreneurs can use these mini case studies on value-chain planning and product design.

The first is a problem that seems to have significant demand, although you wouldn’t know it if you asked – mapping!

People will claim that mapping services will never work because everyone knows every place, because there are street-signs everywhere, and because there just aren’t that many interested people on the net. Its pretty clear, though, that a reliable mapping infrastructure that can connect to other services in the country would have a positive impact on many areas and levels of the economy.

So here are 4 different ways that companies are trying to solve this problem – two of these are startups.

1- Focus on minor steps and build a value chain
I think this is how it started a few years ago – when mapping data was controlled mainly by the Survey of Pakistan, small firms like PRDS with their Digital Atlas of Pakistan (and now Kazakhistan) – these were basically Geographical layers of data over which other firms could add meta layers to build more interesting systems. This data – available a few years ago for $2000 / license or more – was used at that time by telcos, banks and most notably by Honda in their in-car navigation system experiment, which later become a partnership with Trakker, which in turn became the first vehicle tracking system in the country.

Each of these “implementing” groups used the basic data from PRDS, which was the “data organizer” if you will. Over the last year Suparco — our equivalent of Nasa — has also built a completely indigenous GIS mapping and tracking system based on their satellite communication systems. Suparco is also offering their tracking center to tracking services (or just one) in a hosted-type fashion.

In either case, the approach to building maps involves putting a significant staff to the task, which is a long, costly, cumbersome investment. Hence the value-chain – you amortize the total cost of the investment over many smaller licenses that can be offered to a larger customer base. For that to happen, its better if your product is as general as possible and as reconfigurable as possible – in other words, you would focus more on building generic geographical data.

2- Building out a purpose built map yourself
So what did the smaller tracking companies do after Trakker and the #2 firm (CTrack?) got deals that gave them this mapping data in exclusive partnerships? Some of better-backed ones did the crazy attempt to build it themselves.

One of these tracking companies out of lahore actually did something kindof funny. They would dispatch this guy off on a bike and a small device that could read GPS coordinates. This guy would run around all over lahore and stop at every significant landmark, like a bank or ATM or stadium, and then call HQ on a cellphone saying “Oye, I’m here in front of so-and-so, this is my reading”. HQ would then record this off on their map.

Hmm…. reminds me of a certain Verizon commercial, if you ‘hear’ me right. Ha.

Does this work? Maybe the maps are unreliable but its good enough for business, especially if you can sign up customers in long-term contracts. Whats the worst they could do anyway?

3- Nakhsa.net – Scan and convert street maps into vector solutions
The elusive TechLahore (whose identity I have narrowed down to three possibilities) wrote up a bit of a scathing review of Naksha.net on his blog.

Side note: This goes to show – never ask bloggers or the press to look at your incomplete product because first impressions last a lifetime. Stay under the radar until you can floor people with what you have.

Another side note: Usually us tech bloggers wouldn’t even cover incomplete portal-type websites that seem like side-project of bored guys whose core business is something else. I want to cover people who’re betting their entire future livelihood on their products because that commitment usually shows in the attention to detail and quality in the product itself, and it doesn’t wait the time of people who hope the product owners are taking the products seriously as well.

Anyway, the thought process behind Naqsha.net as far as I can tell is that they will take maps that are generally available through the Cantonment Board, and scan it into vector graphic representations. Based on that it seems like they want the site visitors themselves to add the meta data so that it can be built more quickly. The problem is that people in general would not be any good at looking at random lines in a map and guessing what roads they might be. The only end-goal of the content that I can see is more Google Adsense revenue.

In fact, this idea could only work if the experience welcomes participation. I run a company whose core specialty now is experience design, and so if I come across something that asks me to do 4 things before I can actually start using it, I do what Jacob Neilsen predicted I would: I leave!

4- MapMyPakistan.com – A fully community-driven and collaborative map building solution
Umm… *cough*We said so first*cough*

MapMyPakistan.com is a project of 2′scomplement, the small two-person startup started by FAST grads. Although they are not continuing this project any more in the current form, the basic idea is something I have been hoping someone would build for some time now.

Their claim was to let the entire community of residents in a city come to the website and build out the maps themselves – through collaboration, people take ownership in creating useful directions as a group, where each person just builds out a small portion of the overall map.

I’ve been a big believer that this can work, especially by also putting in some local-market relevance into the mix. People in PK dont use street roads to get directions, they use Milestones – ask any person from Karachi or Lahore for directions and they start from the Punjab Chorangi or Sindhi Muslim Circle or Qalma Chowk or some other landmark / milestone.

I was mentoring an entrepreneur with talent to build this out about a year ago but he decided data-mining and building reports was more exciting. I’ve also spoken about this idea to some other entrepreneurs like Monis Rahman, whose team has also looked at hte mapping problem in the past.

It was good to know MapMyPakistan was finally doing it, but sad to see them refocusing into another niche in the area.

Conclusions

So what is the best way of solving this problem? Apparently it seems that the best way is to build so much scale and brand equity that you can influence governments to give you high-quality digital map data. MapMyPakistan has a rumor on their site saying Google is probably going to provide high-end digital maps by 2009. So there.