Andy on Enterprise Software

Pitney Bowes Update

May 8, 2014

At a recent analyst briefing Pitney Bowes executives shared a number of aspects of the current business. The company is an intriguing one, with deep heritage in the US postal business. The company dates back to 1920, and was the inventor of the franking machine. With 16,000 employees it is a major corporation. In more recent times Pitney Bowes has built up a quite large software business, largely through acquisition of both data quality software (Group 1) and GIS software (MapInfo). They do not break out their revenues by business line publicly, but their software business is a lot larger than most people realize. After a period of share underperformance from 2010 to 2012, the stock price has more than doubled since 2012 under its new CEO (Marc Lautenbach joined Pitney Bowes January 2013).

On the software side, Pitney Bowes has had a good reputation for address validation, bolstered by its ability to enrich location data through its full-fledged GIS capabilities, which is a capability that other data quality vendors do not have. However, this potential synergy has not always borne fruit, with a generic sales force not always able to put these value propositions together. The company has a reputation as something of a sleeping giant. A recent change of sales leadership has resulted in changed commission structures and more specialist sales staff being recruited, with some ambitious software sales targets being set for 2014.

A sign that some of this positioning may actually been bearing fruit has been seen in two recent deals in particular, both in the area of geocoding and reverse geocoding i.e. working out exact latitude and longitude from an address, or vice versa. This capability is important in the world of social media as the world increasingly interacts via mobile devices rather than desktops. Recently both Facebook and Twitter have signed major license deals to use the geocoding capabilities of Pitney Bowes. For example when a tweet has a location tag, this uses the Pitney Bowes software. Such high profile endorsements, along with other recent deals such as ones with INRIX (a crowd sourced traffic data company), are potentially very significant. To be able to sign up Facebook and Twitter adds a great deal of credibility to any company claiming location capabilities.

One interesting case study at the conference was that of a UK local authority. A recent boom in construction work in the south east of the UK, but a shortage of affordable accommodation, has resulted in a broad influx of frequently illegal migrant workers from abroad. These building workers are often accommodated in squalid “beds in sheds” at the bottom of residential gardens. These extra residents are not included in official figures and pay no property tax, yet create a burden on local authority services. This particular local authority used reconnaissance planes with infrared cameras to compare human activity at night in the borough with the notional housing. They used land registry data to compare official properties with the number of people actually living in residential areas, often finding considerable quantities of people residing in places where no houses exist, and in these cases avoiding paying local taxes. They were able to use this data, combined via the Pitney Bowes GIS software, to direct police to follow up on concentrations of illegal accommodation, with apparently significant results.

In software terms, the latest GIS release supports 64 bit processing, allowing real time drill down: a demo was shown of zooming from a picture of the earth from orbit directly down to Mount Fuji (as an example) with no redraw delay; this brings the GIS software in this aspect into line with its main competitor ESRI.

Within data quality, Pitney Bowes has street level geocoding support for 122 countries, an unusually deep level of coverage. The company claim that all 25 top US property insurers uses its software. Recently they have added an MDM solution based on a graph database, a logical extension given their considerable customer base in data quality. The unusual database underpinning should in principle allow some interesting analytics e.g. around connections to customer such as which products they use, but also the sphere of influence of a customer, a type of analysis that that a graph database should be able to accommodate more easily than a relational one. However, progress in the market here has been rather pedestrian thus far, and a challenge for 2014 will be to see whether they can develop early MDM sales into a solid cadre of reference customers. Certainly it is always challenging for a new piece of software, even an innovative one, to be marketed against relatively mature and entrenched solutions from large vendors. The coming year will show whether the sales force changes and greater marketing investment around this MDM solution will result in significantly enhanced market share. In our most recent Landscape research, Pitney Bowes customers were amongst the top 7 happiest data quality customers.