How about we take a break from the consumer-marketing world of data and take a little sojourn over to the business-to-business side of Big Data? 

B2B is an often overlooked portion of an overall data strategy. Having an aggressive B2B strategy will help with lead generation, revenue expectation, and an overall understanding of company health. There are fundamental differences between B2C and B2B with the data available and the complexities of collecting it.

The B2B data, while important and extremely valuable, is not as reliable as you are used to seeing in the B2C space. It is also harder to get it to string together accurately.

So, with these difficulties and as a publisher, you are no doubt asking “Why should I make an investment here? Where is the return? And, given it is difficult, should I take the time to develop a strong marketing, data-driven programme for B2B?”

The answer to the questions above is yes.

B2B is different and can be difficult. However, the advantages are huge, once you figure out how to set things up and apply analytics correctly. A typical mid-sized United States market can have 30,000 or so businesses. You have 15 to 20 outside sales representatives.

How can you possibly visit all these people? How do you prioritise, as it is impossible to visit all of the potential customers? The old-style methods of hunting down new business take too long, and it takes a different style of rep to hunt.

Given the resource-to-effort ratio, your operation inadvertently moves into a “safe” mode, and you end up with a sales process that sells to who you’ve sold to in the past and try to squeeze in a little bit of hunting effort. Soon the top level of your sales funnel looks pretty empty.

I propose that you take a little time and gather the data and sales organisations to craft a sustainable and integrated solution: Link your sales and data together. Use data to drive your processes — from CRM (prospecting and lead management) to opportunity and revenue modeling.

If you can use the data and analytics to guide the scarce sales resources to the right businesses with data-driven knowledge, you will easily outperform the random prospecting and reliance on repeat business model that is in place today.

You can leverage the B2B data-driven space with great reliability once you learn, as you did with the consumer data, where the data is directional rather than exact, where your company attracts and where it does not, and, finally, how much revenue potential is available with each company to adjust the sales pitch.

Apply the intelligence to the 30,000 businesses in the market and appropriately target your sales resources. Sure, there are oddities and outright gaping holes in the B2B data, but once you learn where the data landmines sit, they are easily managed.

Finally, as a publisher, if you are not leveraging the B2B space, your efficiency in attracting and building a relationship that works as well as your value and respect in the community as an advertising partner suffers.

This is the case, no matter how good you think you are. Every wrong customer sold or right customer sold the wrong package generates churn and dissatisfaction.

The dream with B2B database building is to connect as many possible sources of business information into one seamless data set.

If you are a traditional print/digital company and rely on selling your own products, it is straightforward. If you are a primarily digital-facing operation and have a lot of retargeted inventory sold or rely on the digital ad networks, the B2B effort is a bit of a twisty road, but it, too, can be navigated.

You have to work hard to find out as much as you can about your site traffic, and connect that B2C behaviour with the businesses performing well on your site.

The goal is to get the right businesses (those categories that end up on your site often) to become direct customers of yours instead of just getting placed on the programmatic/remnant space you set up on your site.

Where do you start?

B2B data gathering, like consumer data, starts with collecting what you already know, consolidating it into a usable form, and then finding the right sources to supplement. The “what you know” data will come from your billing system and CRM tools.

Another internal source, though harder to isolate, is if you have an ad placement self-serve site or a site where a business can find your rate card and product listing. You can discover a lot through the clicks and clickstream data trail they leave behind.

To use a popular cliché: At the end of the day, you need to put together as much information as you can about your current and former advertiser base to begin to develop an advertiser profile.

The next step in building out the profile is to supplement the data you have. You do this by purchasing firm-o-graphic data (demographics about companies) just as you would when building a consumer database. You also buy predictive data such as media spend information.

Unlike the B2C vendors like Acxiom, Experian, and Nielsen, for the B2B data, you will end up with many more vendors to tie everything together. They most likely will not be the same vendors you use for B2C (i.e., build time to lawyer-up for contract negotiations).

There are many companies in the B2B list and predictive space. They each have strengths and weaknesses. Infogroup, D&B, Experian, DatabaseUSA, and Acxiom all offer data, to name a few. I am not in the business of endorsing so I won’t — simply because the “right” one for my needs is different than the “right” one for your needs.

I will say that it is very important that you define the end-goal rather than just pick a company with a well-known name.

If your goal is to build out your B2B database and use it as a source for new, quality, qualified leads, you will end up in a different place (with a different vendor) than if your goal is to understand credit scores.

If your goal is to find businesses that are retail stores versus “all” businesses, you will also find vendor differences. If you are just after a list of names, you will also go in a different direction than if you want to build predictive models from the rich firm-o-graphics some of these companies provide.

My personal selection method for lead list vendors is to set up a data quality/quantity test and see how well your vendor shortlist performs. For example, if you are focusing on retail outlets, give your candidate companies a ZIP code or specific series of addresses (shopping malls or your own address) and see what they have.

Again, all of the suppliers will have missing data. The question is whether the missing data is in an area that is important to you.

The next set of data to purchase requires selecting a vendor in the predictive or market share/spend business. Borrell and Kantar are examples. Again, set up a test. Your state, county, or city should have gross receipts information that you can use to test the validity of the spending projections from the vendors to make your selections.

Once you make your choices, then you can start bringing the data together: your data with the vendor’s information appended and a complete market (or vertical) list.

Appending is where things start to get tricky. Typically your billing system will have the address of where the bill is sent (say, an agency or P.O. Box) while the data source that you so carefully selected has the storefront address. No match! The better providers have several matching methods that they use and combine to lift the match rates. So, consider using them to do the matching rather than a simple in-house matching tool.

Once matched and data is appended, use SIC codes or NAICS codes to group the businesses and spending projections into categories. This isn’t going to be easy. There are more than 1,000 SIC codes in play. A bit of nice “R” cluster modeling from the data analytics folks will get you a segmentation grouping — say around 30 segments — that will work for your market.

From the segments/categories, look at past sales activities of your customers for quantity and dollar amount penetration levels. Don’t just look at the number of businesses as there are some SIC groupings that will never advertise beyond employment ads (most of the pure manufacturing categories).

Some additional work in R, SPSS or SAS — or whatever is your favourite modeling tool is — will let you build out revenue potentials and likely acquisition targets.

A heads up here: When you get into this phase of the B2B space, you will begin to see where the data provider’s data collection and its own modeling will introduce some weird results that you have to noodle through.

(An example: When was the last time you saw a McDonald’s with only two employees and US$80,000 in revenues? This tends to skew how your model behaves).

The strengths and weaknesses of the data in the B2B space make for some interesting data gymnastics challenges that you do not see in the consumer space. These differences mean your data and modeling people are going to have to work a lot smarter to put together meaningful information for the sales departments to use when assigning leads or projecting potential revenues.

But, when used correctly and combined with your sales funnel and world-class sales processes (think e-mail, telemarketing, and in-person), the data-driven lead management will bring an efficiency to your operation as never seen before.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .