Tuesday, June 16, 2009

So you want to start a Geo-Business - Demystifying the Landscape

The foundations of some articles I am currently working on:


It starts with a good idea, moves forward with research, a business plan, maybe financing. Then coding, testing, sales and marketing. Bingo success!

Not so fast. Let's look a little closer.

Historically, the geo-spatial industry has been dominated by a few large companies. The majority focused on geographic information systems (GIS). With the launch of Google maps, back in November 2005, all that changed. The sleeping geo giant awoke. Suddenly slippy tiled base maps were all the rage. Clever programmers learned how to extend Googles map offering, which encouraged Google to make their API's public. Suddenly, we could all easily add maps to our web sites. Customising these maps to include markers for our favorite restaurants, places of business. Overlaying our running, cycling, hiking routes. Yahoo and Microsoft also entered the market. Mapquest extended its offering. ESRI dusted itself off and broadened its business. A revolution was under way. New geo-spatial companies suddenly started appearing. Geo-communities and conferences were enthusiastically launched. Great things were written about the future of the geo-spatial industry; growth, potential avenues of development, the expanding geo-workforce.

Geographers finally felt vindicated. No longer perceived as only 'knowers of capitals of the world'. We had made our mark. Our importance could be demonstrated.

So the geo-revolution is here. Back to our geo-spatial business. Let's plan it here and now. See what we find.

The Idea

So what is our idea? Maybe we should choose something from the current hot topics and refine from there; local search, mobile, data mining.

Let’s consider the current state of play. The common consensus is that, in terms of revenue, GIS amounts to 20% of the market. Consumer mapping applications now dominate.

Let’s presume we have a really great idea (we’ll have one or two later in the article fear not). So have others had the same idea? Time for a little research.

The Research

We need to look at what others are doing. Maybe we can learn from them. Certainly we can see if others are doing what we propose. Define what makes us, or could make us, unique over our competition. Much of this will feed into the business plan. It may also drive us back to the drawing board. Dig deep here. Much can be learned. Time and money saved.

The Business Plan

No need not dwell for too long here. You’ll find a multitude of resources. We will discuss two of the more important sections later. Writing a good business plan is the single most important part of building our geo-spatial business. It is the blueprint. This will guide all work as we move forward. Get advice, don’t rush, be thorough, refine where needed. A poor business plan often leads to a poor business. Start with this:

“Can you describe the business in one sentence and have others understand?”

Making Money

Hmm, making money. Why worry about this now? The ‘build it and they will come’ approach is so much easier. Leading usually to business failure!

This topic should take pride of place in the business plan. In the geo-space, it is a particularly well visited topic. Reviewing opinions and looking at how our (successful) competitors are making money, might prove rewarding.

Three of the most common revenue models are; advertising, charging licensing fees and payment for services.

The Application

Fantastic. We have done our due diligence. A thorough pan is in place. No stone has been left unturned. At last its time to code.

But wait! The geo-technology landscape has broadened. Many choices are now available to us. We need to base our technical decisions on both requirements and cost. Development will be our most expensive pre-launch expense. So think carefully before jumping.

Let’s discuss some of the areas we need to consider.

Rich internet applications (RIA) – Much referenced, but not well understood. Think about the difference between applications which run on a desktop and those which have traditionally run in a browser. Browsers can be thought of as an operating systems, far more limited and limiting than that running your PC. RIA’s bridge this gap. They extend what can be done by browser applications. From asynchronous calls to drag and drop. So what technologies are we talking about here? Best known are those which run in the Flash player - developed in Flash or Flex - and AJAX (a web standard supported by Javascript). So do we wish to build an RIA mapping client?

Open source verses proprietary – We’ve all probably read about the pros and cons. Certainly both have their benefits and drawbacks. Usually discussions center around cost, documentation and support. We need to consider our options carefully, not only in terms of client and server technology, but data. For GIS do we go with the ESRI ArcGIS stack or maybe look for a pure open source option? Maybe even a combination. Consumer maps could be built using the Google, Microsoft, Mapquest or Yahoo API’s or possibly Modest Maps.

The Data – We have no application without good data. But what form does that data take, where is it stored and what is the source? GIS applications usually rely on shapefiles and geodatabases. Mash ups will use a variety of sources RSS/GeoRSS, maybe RESTful calls to Flickr or Trulia. Is the data free? Do we need to gather and generate it? Where will we get our base map?

Mobile verses Web – with the launch of the IPhone, mobile application development has taken off. Most mobiles now have built in GPS. Thus the potential for mapping applications is considerable. Should we also be considering mobile as a potential platform for our application? Will our mobile application mirror our web application, or be complementary?

At this point we could look in more depth at the many options possible from the above discussion. To avoid this potentially long laundry list, see the following web page:


This walks you through some of the options.

Getting Noticed

We have arrived. Under budget and proud of our, oh so useful mapping application. Now for the exciting launch and promotion. Back to our business plan, under promotion an empty space! One hopes not. Without great marketing good ideas die and great products disappear. Let’s cover briefly some of the options. Traditional promotion remains important, but here we will focus mostly on online advertising. We can break things down as follows:

Search Engine Optimization (SEO)

“your goal is to rank for free within the organic search results for target keywords related to your Web site”

Ranking means we appear high up in a search engines listings, after a relevant keyword search. Put simply SEO is about business relevant web page content, good metadata and back links. Time need be spent generating not only a memorable business web site, but one that is optimized. Choosing appropriate keywords is a critical part of this process. There are a range of good tools to measure SEO. Google provide analytics and a toolbar, which includes the all important PageRank. There are also a variety of useful browser plug- ins including SEO for Firefox. Many write their own tools, tracking for example server and database statistics.

Social Media Optimization (SMO)

a form of marketing that focuses on generating traffic and buzz by participation in various social media Web sites”

SMO is really an extension of SEO. It involves joining social networking communities. The most popular sites are as follows:

Facebook - interact with current and prospective business associates.

MySpace – a vehicle to manage reputation, build brand recognition and promote products.

StumbleUpon – includes a toolbar to discover and rate web pages, videos, photos. Getting your Web assets submitted to StumbleUpon is an effective way to generate traffic, and build backlinks.

MyBlogLog - installed on your blog or content Web site. Shows your visitors the most recent MyBlogLog members who have accessed your Web site.

Twitter – micro-blogging or ubiquitous text messaging.

LinkedIn – business-oriented social networking

Del.icio.us - a popular social bookmarking Web site. Allowing users to browse the bookmarks of other people with similar tastes

YouTube – Video posting service. Uploading interesting videos here can help share company information and generate interest.

Digg – news stories are submitted, then voted on, and maybe promoted to the front page.

Phinn – similar to Digg but aimed at interactive marketing professionals.

Search Engine Marketing (SEM)

“your goal is to pay for placement by competing with other advertisers for ranking within the sponsored results section of search results.”

SEM includes pay per click (PPC) campaigns. Many suggest this is a vital part of any online promotional effort. Google AdWords is the largest and most popular PPC advertising platform. How it works is that you bid on a group of carefully chosen words. If you are successful you then generate an ad, which is carefully worded using your keywords. When a user conducts a search using any of these keywords, your ad appears on the page. Each time a user clicks the ad you pay Google the fee agreed in the bid. You can target your ads in various ways. The most common form is geo-targeting, which relates nicely to our discussion. One more point. Make sure the landing page from an ad click is well constructed, contains our keywords and the text relates to the ad copy.

Other areas to consider under SEM are indexed images in Google Images and optimizing your site for the EBay and Craiglist search engines.

Additional Promotional Tools


Blogs are a great way to share announcements, technical information, thoughts and news. Most popular blogging tools are Googles Blogger and WordPress. The goal of our company blog is to generate content that is so interesting and useful that others link to our Web site. Remember to always write search engine optimized posts.

Technorati is a popular search engine for searching blogs. If you can get your blog to rank well on Technorati, you can gain considerable exposure and traffic.


There are an increasing number of international geo-spatial conferences. Where relevant, you should attend as many as possible. They are great way to network, stay up to date on new developments and assess the competition. Maybe present, set up a booth, sponsor an event.


We touched on only a few key promotional tools. There are many others; starting online forums, writing articles, to name but a few.

Pulling It all Together

Let’s be bold. From our discussion let’s discuss starting that geo-business. For the sake of our discussion we will talk about two potential businesses; GeoCommunities, a local data consumer mapping offering and the GIS based MapMyWalk. In each case we will presume a very limited budget.


Our focus; the consumer market. We are thinking in terms of a local data mash up.

Our competition includes Mapquest, Google, Yahoo on a big scale. And smaller companies like Everyblock. So how are they making money and what makes our offering unqiue?

Everyblock is in an interesting position. They have built a hyperlocal search site providing news and other relevant local data. Their revenue model is still be decided, but likely to be ad based. Their code base will open sourced in June 2009. A potentially challenging event. The search based companies use purely ad driven revenue models. They include ads within their search page. Thus opening local search in Mapquest will show ads intermingled with search content. Increasingly these companies are looking to make ads relevant to a users search. Thus a restaurant search in San Jose will also show ads for local ATMs and movie theaters. Maybe we should look more closely here.

Let’s focus on our own local market. The place we know best. Can we provide data that will be useful to the local community? Maybe data we gather, or that out of the reach of the big search engines. We might encourage our growing user base to provide data on local events, recommendations, concerns. Let’s go with an ad based revenue model for our business. But focus on local ads combined with local content. Thus a search for preschools, will map all in the defined area, and include ads for select paid preschool advertisers.

We want the application to be as user friendly as possible. A rich Internet mapping application is our first choice. Thus we will be using Adobe Flex 4 to build our user interface, purely based on familiarity. Given our budget we favor open source. The Flex SDK is open source, though the IDE is not. Modest Maps will be our Flex mapping library. For our base maps we will use roads data provided by Open StreetMap.

It’s worth mentioning the API’s provided by the likes of Google. These provide excellent tools, making development quick and easy. There are two problems though. Their API is locked. Meaning customization is impossible. Also, and of greater significance, their (free) data may at some future date carry advertising. Thus zooming to a point of interest may also reveal markers for ‘the best pizza joint in town’! Potentially a big turn off for users.

Much of our overlay data will be pulled from RSS/GeoRSS feeds and RESTful calls. But we will need to store and update data in a database. Again turning to open source, we will use MySQL. There are other options, notably Postgres. Since we will be using PHP in the server layer, running on Apache, this integrates well with MySQL. PHP will both help overcome some of the security issues we encounter with the Flash player, and will interact with the database.

So the coding begins. A test application is built:


Note, a tutorial on how to start building a mapping application using the architecture proposed here is available via the Adobe Flex Developers Center or the flexmappers blog.


Our second company will be built using GIS. The company specializes in route mapping, but aimed at pedestrians. For example, a tourist visiting an unfamiliar city, wishing to see some of the historical city buildings. Our service will allow users to either map a route themselves or select from a list. Points of interest (POI) on routes will be included. Routes stored in our web application will also be available via a mobile device.

How will we make money? We will use a subscription model. Thus, when visiting our Web site, users will be able to create routes and view a subset of predefined routes for free. A subscription will give them every predefined route for available cities and, most importantly, the ability to save ‘myroutes’. Accessing these routes via their mobiles.

Let’s look at our competition. Schmap are doing something similar to what we are proposing. So how will we be unique? In a number of ways. First we will actually map (shortest) routes and not just simply add map markers. Second, our UI will be richer and more interactive. Third, we will use a highly detailed base map.

Let’s jump ahead. So we’ve written our comprehensive business plan. Included a thorough promotional campaign. It’s time to think about the application. The ESRI ArcGIS stack is certainly an option we should seriously consider. ESRI have written excellent RIA API’s and ArcGIS can handle all our routing needs. But, since we are tight on budget, we will first look at our open source options. One of the more popular open source GIS stacks has been constructed using Openlayers, Geoserver, GeoWebCache, PostGIS and Postgres. Data is a key consideration. One of the factors which makes us stand out is our detailed base maps. It’s likely we will need to buy this data. Companies like UnitedMaps provide data with this level of detail. We will need to budget for this expense.

There is a very active community using the various open source components in this stack. A number of the core developers for GeoServer and PostGIS work for a group called OpenGeo, and provide excellent support. We will need to work a little harder to get our stack capable of providing our routing needs. For example, the use of the pg_routing extension for Postgres.

We would need another article to walk through these development steps in more detail. One last thing to add. Mapping options on mobiles remain somewhat limited. Phonegap is an open source project of potential use. The Actionscript 2.0 library of Modest Maps can be developed in Flash Lite and runs on any phone which has the Flash player installed. Lastly, Nokia are about to release a mapping component within OVI. This may also prove useful, as long as we can access our server components.

Promotion, of course, will be crucial to the success of both companies. After all, we won’t have a highly paid sales team. A key component will be attracting visitors to out web site. That’s where our carefully planned SEO, SMO and SEM campaigns will be essential.


This article has walked you through setting up a geo-business. The market has broadened. Offering opportunity and increased complexity. Hopefully, this discussion has helped demystify the current landscape.



No comments: