The Global 5000 database is an effort to create and maintain a database of the 5,000 largest (as defined by annual revenue) companies in the world. These can be public or private firms and are located in any country. It is built by hand with a great deal of research depth and updated on an ongoing basis.

Inspiration for building The Global 5000 was born out of two factors that happened about the same time:

  • For a research project, I needed to develop a market map of revenue by industry for the top 2500 companies in 12 major countries
  • A client had asked for help building a database of the 1,000 largest private companies in the world

While we could find some existing products that would fulfil some of the requirements, there was nothing available that fit exactly. We had to customize existing data in some way. After a lot of effort, it became clear that we had to build & refine on top of existing data. In the end, it would have been better and quicker to build it ourselves

Building The Global 5000 database was not an idea that sprung up overnight. I had the experience of working with other B2B databases – in the early days of ComputerWorld and IDC followed by managing the CI Technology database under Ziff Davis prior to its sale to Harte-Hanks. I knew what I was getting into. Now, after a few years of running this database, I am happy to share some of the learnings. For others who want to venture down a similar path.

  •  It is a lot easier to build it than to keep it updated.

Creating the database is an easy first step. There are a lot of resources available online and internal company data is often readily available as well. So, the resources are there to start. But to keep it relevant, there has to be a dedicated, continuous effort to update information. Companies merge, sell off divisions, file for bankruptcy, spin-off product lines, move locations and any number of other activities on a regular basis. To keep up, you have to stay on it every day, week and month. Don’t think you can easily re-visit every quarter or year. Might as well start from scratch.

  • Use company/organization data as your basis. Attach/link people info to that company record

In many sales & marketing applications, the natural tendency is to think about people first and build the data around the contact info. But people change more frequently than company info does and after that individual leaves … you will still want to market to the company.

For updating purposes, you can likely update company info on a 12-18 month basis as a routine. For people info, a 6 to 9 month interval is best to catch all the changes.

  • Put someone in charge who will has a detail, data slogging mindset but can make reasoned judgment decisions

Much of the work involved will be drudgery and rote but putting it in the hands of the inexperienced is not the best solution. If the database is large enough, using machine based tools is the way to go. Someone needs to be the owner of the rules to set up algorithms and put in place a qc process.

Don’t assume the IT/tech/developers will own the data. They typically don’t care about the data. They care about where & how it is housed, getting you access, getting it out but not worried about the freshness of the data

  • Too much data early on can sink the ship

For all the talk of Big Data, go slow in the beginning. Do you really need to have every instance of a company mentioned be adding to & updating the records in your database?  Probably not. There is a tendency to try and think of & include everything from day 1. If you did that, you probably can’t keep up

  • Decide if you are going to keep the data at a HQ level or location or a business entity leve

You have a lot of choices and they should be based on where you sell. Who owns the buying decision and the budget? Some products & services are sold at the local, individual location level – some are sold to HQ. This becomes particularly important if you are going to buy data from any of the data suppliers. Know how they have organized & how they sell their data

  • Start

If you wait for every decision to be made and vetted within the organization, you’ll be months or years before getting off the ground. Put the data into something simple to start getting your mind around processes, where to add update dates, what other fields are needed. Put it in Excel – it can be imported from there into anything you will build

  •  Determine what is good enough

For those who are perfectionists, this project will drive you nuts. The data is never done. Updates are always needed as the external world continually changes. Researching every last item and trying to keep it updated is a never-ending task an d organizations need to understand where they have ‘good enough’ data to use for 90% of the applications. You can purchase good enough and add specific twists for the company needs, but don’t expect to get it 100% right and keep it that way going forward.