Memorandum by Getmapping
I am responding on behalf of Getmapping plc
to your letter of 24 April 2007 to Sir Bryan Carsberg asking for
submissions for your committee's forthcoming review of Ordnance
Survey (OS). Getmapping is a member of the Locus Association.
Getmapping is an OS Licensed Partner and a reseller
of OS's mapping products. However, Getmapping's main area of business
is to create and maintain a national aerial photographic survey
of Britain (The Millennium Map) which is in direct competition
with OS's Imagery Layer. Thus our main concerns with OS are from
the point of view of a direct competitor and not a re-user of
OS data.
Before answering your specific questions, I
would like to try to give you a clear idea of how difficult Getmapping
is finding it to operate as a competitor of OS now that OS is
acting as an unregulated commercial entity with unclear boundaries
between its public service and commercial activities. We think
that these difficulties, experienced by many companies in the
Geographic Information (GI) field, have badly damaged the development
of Britain's GI industry.
I have also attached a number of Appendices
to illustrate some of the issues raised in this letter and to
demonstrate that the GI industry has been unable to find a satisfactory
channel through which to have them addressed. Since some of the
Appendices were sent in confidence I would be grateful if at least
the Appendices are not made publicly available.
A BRIEF HISTORY
When we launched Getmapping (then the Millennium
Mapping Company) in 1999 OS was not acquiring any colour photography
and was not producing any mosaiced (ie seamless, map-accurate)
aerial photography products. Furthermore OS had no plans to do
so. OS therefore became a reseller of Getmapping's Millennium
Map and the two organisations seemed set to work together in a
long term and mutually-beneficial partnership. Had this state
of affairs continued I believe that Getmapping would by now be
a much larger and more profitable company operating on an international
basis. We were the first company world-wide to develop a private-sector
national aerial survey served via the internet, and our plan was
to develop an international business from a solid UK base.
However in 2001 OS decided to compete with Getmapping
by creating and owning its own Imagery Layer, an equivalent of
Getmapping's Millennium Map. In Getmapping's view OS's Imagery
Layer was going to be subsidised from public funds and so in 2002
Getmapping took legal action against OS to try to prevent OS from
competing unfairly against Getmapping in the imagery market. Getmapping
had to withdraw from this action before it came to trial because
the company had insufficient funds to continue, and Getmapping
also lost its appeal for an injunction to force OS to include
the Millennium Map within OS's Digital National Framework. The
judgement from the injunction hearing was highly critical of Getmapping's
case, and Getmapping was ordered to pay OS's costs which were
over £450,000 (although Getmapping and OS eventually agreed
on the payment of a smaller six figure sum which our Settlement
Agreement prevents me from disclosing). See Appendix A for a description
of the issues behind the case.
Getmapping still believes that OS is competing
unfairly in the imagery market. We think that OS can do this because
the boundaries between its national interest and its commercial
activities are not well defined and because there is no effective
regulator to whom we can appeal for help. Had OS taken notice
of the recommendations of the previous 2002 Select Committee report
then we think that many of our current problems would have been
avoided.
COMPETING WITH
A TRADING
FUND: A COMPETITOR'S
VIEW
From Getmapping's point of view, competing with
the OS in the imagery market has been a dangerous, frustrating
and uncertain process.
— From 1999-2001 we worked in partnership
with OS with the expectation that OS would resell the Millennium
Map and use it as an Associated Data layer within OS's Digital
National Framework.
— From 2001-03 OS was a direct competitor
of Getmapping and OS's intention at that time was to acquire colour
imagery only once (for map update and the Imagery Layer) and to
offset the majority of the Imagery Layer costs against its cartography
budgets. In reaction to this situation Getmapping developed a
new strategy which involved abandoning national maintenance of
the Millennium Map. Instead Getmapping would become a licensed
reseller of OS's Imagery Layer and Getmapping would supplement
the Imagery Layer with its own higher resolution photography of
key areas of the country. However, it soon became clear that OS
did not have the power to license its Imagery Layer to its Licensed
Partners.
— From 2004-06 it became apparent,
although nobody ever told us, that in fact OS was not after all
permitted to re-use its cartographic photography within the Imagery
Layer. OS was in effect having to acquire (or procure from third
parties) the photography twice, once for cartography update and
once for the Imagery Layer. However, during this period OS explained
that from 2007 they would return to the strategy of flying the
country once for both cartographic update and the Imagery Layer.
Suppliers of photography would not be allowed to retain ownership
of the photography, as had happened for the first layer of the
Imagery Layer. OS justified this policy, which in our view amounts
to a cross-subsidy of funds from OS's dominant and unassailable
position in the cartography market, on the basis that it was now
winning its cartography contracts for Local and Central Government
under open competition (the Mapping Services Agreement (MSA) and
Pan Government Agreement (PGA) respectively).
— 2007. There has been no sign of any
procurement activity for OS's second layer of the Imagery Layer,
and we have no idea of their current intentions. However we have
recently discovered that OS is already using some of its cartographic
update photography in the Imagery Layer which in our view is a
cross-subsidy from its public service to its commercial activities.
It has been very difficult to develop the company
in the face of these swings in OS's intentions—we know that
OS intends to make our current business model unsustainable, and
we have to spend a lot of time and effort planning alternative
business models and trying to find a regulator/arbitrator who
can help us.
Throughout this period of uncertainty Getmapping
has tried to establish a channel through which it could voice
its concerns and have them debated and resolved openly and fairly.
We have tried OS itself (at all levels), DTLR/ODPM/DCLG (at all
levels), OFT, HMSO and the National Audit Office. In 2003 we founded
the Geographic Information Forum (GIF) to try to get the weight
of several companies behind the need for recognition and resolution
of the problems. And in 2005 we were closely involved in the establishment
of the Locus Association. We have used our own resources as well
as help from our MP. Through the GIF and Locus we have engaged
in political lobbying in Britain as well as lobbying to the European
Union. See Appendices B.1 to B.4, C.1 to C.3 and D for some examples
of our attempts to try to establish a channel for complaint: none
of these efforts has so far resulted in any useful advice or support.
From the point of view of a competitor of OS
Getmapping has formed some strong impressions about OS, its regulation
and the effect of the current situation on our ability to compete
effectively.
(a) Competition versus Co-operation
As a Trading Fund OS has a responsibility to
act as a commercial entity and make a profit. This is bound to
clash with OS's remit to act as a force to develop the GI Market
in Britain for the good of all.
We think that OS has not paid nearly enough
attention to the effect of its actions on other companies within
the sector. OS is in a dominant position in the market and it
is (in effect) state funded (through un-competed or ineffectively-competed
mapping contracts for government and the utilities). From this
dominant position OS ought to have a duty to ensure that it deals
fairly with the rest of the market. In our view it does not do
so.
(b) Regulation/Arbitration
As far as we can see there is no effective regulation
of OS. None of Getmapping's complaints has seemed to result in
a useful outcome. Furthermore I personally believe that in the
case of Getmapping's 2003 complaint to HMSO, Geoff Sadler (who
was responsible for our case) was not given the freedom to deal
with it as he saw fit. I believe this because at our first meeting
Geoff Sadler was very sympathetic and said that there did appear
to be an abuse. He said that at the very least he would ensure
that we were given a fair hearing and that he would get the parties
together to discuss the issues. However, Getmapping's complaints
were dismissed. HMSO's report appeared to contain a fundamental
discrepancy. HMSO acknowledged that OS "should not use its
market power to compete unfairly"" but then said that
"where there is mixed use of imagery for cartography and
the Imagery Layer, the costs are apportioned. HMSO are satisfied
with this approach." We asked for a justification of this
discrepancy but we were never able to discuss this with Geoff
or anybody else. See Appendices C.1 to C.3 for details of the
complaint, HMSO's response and Getmapping's counter-response.
The regulatory framework appeared to be moving
in the right direction with the creation of OPSI, and we were
all encouraged when in 2006 Intelligent Addressing had a complaint
upheld by OPSI. However, it now appears that OPSI has no power
to enforce its recommendations, and so we are no further forward.
This situation is analogous to a dog (OS) worrying
sheep (OS's competitors). From 2001 to 2006 the sheep bleated
but there was no owner (regulator) within earshot. When at last
the owner listened (OPSI in 2006) and ordered the dog to heel,
the dog ignored the command. We think that the dog must be put
on a lead (through effective regulation) and/or the sheep must
be enclosed within a field (ie there must be clear boundaries
to the scope of OS activities). Only when the sheep are safe can
they develop properly and attract investment.
Despite the apparent lack of success, we suspect
that behind the scenes things may not be quite as bleak as they
seem. For instance, although Getmapping lost its injunction hearing,
and although its complaint to HMSO was dismissed, OS for some
reason never carried out its stated intention of using the bulk
of its cartographic photography to populate its Imagery Layer.
Perhaps after all there is a mechanism, invisible to the private
sector, by which OS is to some extent controlled. However, we
think that it is more likely that OS is aware that it cannot afford
to be too aggressive in its treatment of its competitors for fear
of arousing the interest of the regulators.
(c) Consultation and Communication
Since 2000 OS has stopped publishing the Information
Papers and Consultation Papers which used to keep the industry
informed of its plans and intentions. OS has never consulted Getmapping
about its plans for the Imagery Layer or listened sympathetically
to Getmapping's concerns about the effect of its plans. At a working
"Account Manager" level OS staff have been helpful and
co-operative. However, the Account Managers seem ignorant of,
and are certainly not able to influence, OS's policy. At the policy/director
level OS's attitude to Getmapping has generally been aggressive
and arrogant. Several times Getmapping has been urged to become
a supplier of imagery to OS and to give up its aspirations of
owning its own data.
(d) OS Remit and Boundaries
It seems to us that OS has at least one market
in which it is both dominant and in effect publicly funded. This
"national interest" market is the creation and maintenance
of national large-scale mapping (MasterMap). However, there are
other "commercial" markets in which OS is not in a dominant
and publicly-funded position. Aerial photography is one such market
in which the private sector is stronger than OS, and so far the
OS Imagery Layer has not made very much impact on our sales. In
these markets OS should compete on equal terms with the private
sector.
We think it is absolutely vital that a clear
boundary is drawn between OS's "national interest" markets
and its "commercial" markets, and that procedures are
put in place to ensure that there is no form of cross-subsidy
between one and the other. At the moment there are no clear divisions
between the two types of market, and there are no clear rules
to prevent cross-subsidy.
(e) Unfair Practices (with an Example)
In Getmapping's opinion, OS indulges in practices
which are blatantly unfair and which it knows to be unfair. OS
acts unfairly because it calculates that it can get away with
it and because it puts its commercial interests before the wider
interests of the GI industry. OS is encouraged in this behaviour
by the knowledge that there is no effective regulator.
I will provide a single current example to illustrate
this point. OS is competing with Getmapping for the provision
of the Imagery Layer, and from 2004-06 OS has been funding its
Imagery Layer photography separately from the photography it uses
for updating its cartography. In other words OS is effectively
funding two layers of photography, one for the Imagery Layer and
one for cartographic update. This is good for Getmapping because
there is no cross-subsidy of the Imagery Layer from OS's cartographic
budgets. However in June 2005 OS announced to Getmapping that
from 2007 it would procure a single layer of photography that
would be used for both the Imagery Layer and the Cartography update
programme, and that OS would offset the cost of the photography
against its cartographic budgets. OS's justification for doing
this is that all the government's mapping requirements have now
been competed (under the Mapping Services Agreement (MSA) for
Local Authorities in 2004 and under the Pan Government Agreement
(PGA) in 2006). Because OS would win this mapping work in open
competition then OS was justified in arguing that it was no longer
in a monopoly position and that it could therefore offset its
Imagery Layer costs against its cartography programmes.
If OS was allowed to do this then Getmapping's
current business model would be unsustainable. We would have to
pay (about £1 million pa) to acquire our photography while
OS could offset those costs. Getmapping simply could not compete.
OS's argument is clearly unfair because both
the MSA and PGA competitions were designed in such a way that
only OS could supply the cartography. (They could have been designed
differently, but ODPM, of which the Director General of OS was
a director, chose not to do so). How can we assert that only OS
could win? The main reason is that the winning contractor was
required to supply complete large-scale cartography for the entire
country within 6 weeks of winning the contract. This is clearly
impossible unless a bidder was prepared to invest in creating
an entire large-scale map of Britain in advance. This would take
about five years and cost perhaps £100m. Please see Appendix
F for more details of why we believe the PGA procurement was unfair.
We believe that OS can only indulge in such
blatantly unfair practices because it knows that there is no effective
regulator and that its private sector competitors cannot risk
taking them to court.
(f) Effect on Getmapping's Ability to Develop
its Business
Getmapping entered the Imagery Layer market
in 1999, and we floated the company in 2000 with the aim of developing
an international imagery layer which would be served online. In
other words we wanted to create a Google Earth. Getmapping was
the first company in the world to create a privately-owned national
scale aerial survey.
Until 2001 the business developed well in the
UK. In addition we launched a Spanish and an Australian joint
venture and we opened discussions with third party data suppliers
in many other countries (Sweden, Germany, France, Holland etc).
However when OS announced its plans for its
own Imagery Layer in May 2001, the whole picture changed. Our
sales were damaged by the loss of OS as a reseller; public sector
sales were damaged because OS failed to announce any prices for
its Imagery Layer; we had to abandon our 50 cm survey of Scotland
because OS offered us a contract to license our Scottish data
from us provided we re-flew it at 25 cm; our shares crashed; and
we spent nearly two years and c £300,000 on legal fees to
try to prevent OS from abusing its dominant position. As a result
of this the company nearly failed. We had to cut our staff, abandon
our overseas plans and retrench.
However, the effects of OS as an unregulated
Trading Fund go much further than that. Even now we find it very
difficult to plan our future when we don't know whether OS will
be allowed to offset the costs of its post 2007 imagery against
its cartography budgets. If it is allowed to do this then Getmapping's
current business will become unsustainable. Our difficulty is
that we cannot raise money for new investments or make a proper
business plan until we know how these issues are going to be resolved.
Every year we have to reflect the uncertainty over OS in our Annual
Report and Accounts, and this keeps our share price depressed
and makes it impossible for us to raise more money for investment.
THE COMMITTEE'S
SIX QUESTIONS
Many of the Committee's six questions have already
been covered to some extent in the paragraphs above. Additional
comments are provided below:
Q1 Boundaries of OS's Public Service and National
Interest Work
We believe that very little has changed since
2002. We think that it is vital to define clear boundaries between
OS's National Interest (and quasi-publicly-funded) activities
and its Commercial Activities and to ensure that there is no cross
-subsidy between the two.
It might also be helpful to define a third category
of activity in which OS will not be allowed to operate at all,
for instance publishing road atlases. This would give companies
a clear indication of where they could build a business without
fear of competition from a dominant National Mapping Agency.
Once defined we believe that the boundaries
between these three classes of activity should not be allowed
to change without extensive consultation and a five year warning
so that companies involved can make plans to adjust their business
plans accordingly.
Q2 Effective Arbitration
We think that the current arbitration process
is hopelessly weak and inadequate. Furthermore we do not believe
that it is independent: we think that HMSO was prevented from
dealing properly with our 2003 complaint.
However, it does appear that behind the scenes
OS may sometimes be constrained by somebody. An example of this
is that, despite saying in 2002 that it would replace the first
licensed layer of the Imagery Layer with its own cartographic
photography, OS has not done so. This is a mystery to us.
We would also like to stress that an effective
arbitration system must allow for a complaint to be made in a
simple letter and in layman's terms. The onus should then be on
the arbitrators to get to the bottom of the issues and make their
ruling. If a company has to spend perhaps £50,000 paying
lawyers to formulate a complaint then most issues will never see
the light of day.
Q3 What is your Assessment of UK Geographic
Panel's operation?
Getmapping has very little knowledge of how
the panel is working. We have had no contact with the panel and
have not seen any significant outputs.
However, we think it is wrong on two counts
that OS leads the panel that provides strategic GI advice to the
government. First, OS now has a commercial interest, and we believe
that OS would naturally use its position on the GI Panel to further
that interest. Second, OS is a supplier of maps, and it will naturally
lean towards a Rolls Royce solution to mapping products which
may be far removed from what the users need. For example many
users consider that MasterMap is over-specified and too sophisticated
for the majority of its users. And yet OS was able to specify
it on behalf of government and the utilities who are now forced
to pay for it. We think that it would be better for the specification
of the customers' needs to be entirely separate from the suppliers.
Q4 Is the UK Geographic Panel's membership
Well Balanced?
No, it is far too biased towards OS and the
public sector. In addition none of the GI panel members has ever
made an approach to Getmapping to ascertain our views about National
GI policy.
Q5 Should DGOS advise the Government on GI?
No, DGOS should definitely not advise the Government
on GI issues. DGOS is almost bound to try to influence policy
to benefit OS's commercial position.
This would apply even if there was a clear separation
between public service and commercial activities because DGOS
would still have an interest in providing advice which benefited
OS's Commercial arm.
Q6 Unfair Competition
A lot of the Q1 arguments apply here. OS is
unquestionably in a dominant position in the mapping market—this
dominance is perpetuated by the MSA and PGA competition structures
which do not make it possible for anybody to compete for the large
scale mapping—therefore OS receives over £50 million
of public money per annum to maintain the country's mapping—OS
sees no distinction between its public good and commercial activities
and therefore feels free to cross-fund between these activities.
This is still the case—OS is now embarking of a policy of
"fly once and use may times" for aerial photography
which means that OS still intends that its cartography budgets
will underwrite its commercial Imagery Layer.
This takes us back to Question 1. The solution
is to define clearly each of OS's activities as "Public Good",
"Commercial" or "Out of Scope". Then there
needs to be a clear procedure to ensure that there is transparency
and no subsidy between the Public Service and the Commercial activities.
SUMMARY
In summary, Getmapping is in the business of
providing a product which it has to pay for itself. Getmapping
competes against OS which is allowed to offset the costs of its
photography against its Public Service cartography budgets. This
behaviour has been specifically condoned by HMSO. So far OS has,
in the main, chosen not to cross-subsidise the bulk of its Imagery
Layer. However, from 2007 it intends to do so. So Getmapping is
in effect living on borrowed time—it should not be possible
for Getmapping to compete effectively with OS under these conditions.
This is both unfair and bad for the development of the GI industry.
In 2002 the DTLR Committee correctly identified
the main problems with the structure of OS and made clear recommendations
for putting them right. All but one of the recommendations were
ignored by ODPM/OS, and as a result Getmapping and the other direct
competitors are still suffering from an unfair competitive environment.
Getmapping hopes that the Committee will recognize
the problems that still exist and that it will re-iterate its
recommendations. Furthermore, Getmapping hopes that the committee
will consider mechanisms for ensuring that its recommendations
cannot be ignored for a second time.
| 
