« Home »

The Seamless Audit Trail

Eric Cohen and other XBRL-GL evangelists have been advocating the idea of a seamless audit trail for some time now. Potentially there’s a lot of complexity to this idea involving metamodels of ERP data in UML and so on. But I’m a simple person, so I tend to think of the seamless audit trail as a refinement of a basic function of any accounting or financial reporting software package: Drilldown.

Drilldown is a commonplace feature of all accounting/ERP/financial reporting software. The aim is to enable you to get from summary to source i.e. to navigate from information to data or from report summary line to transaction posting line. The problem is that every transaction management software package is different so it’s not easy for any reporting package that sits above these to easily understand how to navigate down to the transaction data source without maintaining multiple ‘mapping layers’ into various ERP systems.

XBRL does a great job of standardizing ‘top-level’ report formats to provide an ideal start-point for a drilldown. But without similar taxonomy based tagging at ledger account and perhaps even transaction level line level, it’s hard to create a ‘universal’ drilldown capability that can facilitate a seamless audit trail. And it’s even harder when the source data systems are not part of an integrated ERP suite but spread all over the place as stand alone modules on an organization’s LAN or WAN.

That’s why it’s a shame that ERP vendors have not embraced XML-based web services as well as they might have. Every module in an ERP system essentially functions as both a data receiver and provider  - that’s why traditional data import and export functions are important core functionality in any ERP system. And enabling data consumption and provision programmatically across the Internet is basically the point of web services.

So one way of providing a seamless audit trail, top-down from an XBRL formatted report, is to be able to call a standardized GL/AR/AP etc. web service (or data interface) and pass XBRL tags as a means of navigating the drilldown hierarchy . Of course this also requires XBRL tagging to move down and become embedded at least at sub-ledger account level. Which is one reason why the first stage of XBRL-enabling ERP systems for seamless drilldown, XBRL-GL, is an important step towards this goal.

ERP vendors seem to be genetically pre-disposed to avoiding any kind of standardization that may reduce their competitive differentiation. But here we are just talking about a very simple, but standardized, ‘GET’ interface available for each ledger that would also facilitate ERP-to-ERP interoperability generally. Unfortunately without this kind of joined-up thinking it’s not just the seamless audit trail that remains pie-in-the-sky, it’s also the promise that XBRL holds for improving the control of internal accounting data and re-architecting the way consolidation reporting software functions.


Tags: ,


This entry was posted on Monday, January 25th, 2010 at 4:02 am and is filed under Control, XBRL. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

  • Pingback: Tweets that mention The Seamless Audit Trail | Rivet Software -- Topsy.com

  • Daniel Roberts

    Stewart,

    The ERP vendors have invested heavily in seamless audit, and will happily help clients implement – for a price. This investment required a strong understanding of their complete infrastructures of which data structures are only one part. Changing the data structure would be very expensive and time consuming, and a strong business case would be required.

    Unfortunately the "killer" section of the business case has a section header of "Competitive Threat" and discusses the potential for lost business if their products become fully open and independent standards based.

    The good news is that for some, open standards will eventually prevail. This will be at the "lower end" of the ERP market, where providers will gain a competitive advantage by being open, or non-proprietary.

    The historical analogy is the computer operating system environment. twenty five years ago there are a number of operating systems, from DOS, to MVS, VM, Prime, Wang, DEC, Sun, DataGeneral, ICL, Honeywell. The list goes on. And into this mix came Unix. And Unix was small, and "open" and shunned by all of the very large systems providers. Today there remain a number of operating systems, but almost all will run an Unix shell to allow any Unix application or environment to run seamlessly.

    But the move to Unix almost killed the computer majors, or perhaps they were almost killed by their lack of flexibility in responding to the market.

    Unfortunately I do not yet see the Oracles, SAPs, etc, cringing in fear of open sources standards for either data or accounting / ERP systems. When that happens, we will see the business case change from "Competitive Threat" to "Competitive Opportunity". Then and only then will the investment be seriously considered.

    • http://www.rivetsoftware.com Stewart McKie

      Daniel

      Thanks for commenting.

      I have some reflections on your points.

      1. I don't see much evidence that ERP vendors have invested heavily in seamless audit. They were very slow in adopting drilldown effectively in the 90's, let alone seamless audit in the 00's. But I am happy to be corrected.

      2. The whole point of XBRL is that it doesn't substantially change existing data models at all – or doesn't need to – XBRL is metadata. To tag an account does not require any change to an existing data model except an additional column in a table (I'm slightly oversimplifying of course). So retrofitting XBRL is hardly rocket science.

      3. As I said, competitive differentiation is an issue for resisting open standards. SAP didn't get to where it is by being open. So I agree with your point about 'competitive threat'. I also agree that the impetus for truly 'open accounting' is likely to come from the bottom-up rather than the top down. But the ERP marketspace is notoriously difficult to shake-up even when XBRL presents an opportunity for a new generation of 're-designed from the ground-up' accounting software.

      4. I'm not sure I understand what your point is about UNIX. It certainly impacted lots of hardware vendors (e.g. DEC) but had little impact on accounting software as many vendors thrived in the 80's and 90's without having UNIX versions. In the 1980's I sold some of the first UNIX accounting systems in the UK on some of the first 'name brand' UNIX boxes (does anyone remember the IBM RS600 or the AT&T 3B2?). The only reason customers bought these was nothing to do with open standards but because they could run a multi-user accounting system without having to buy a VAX or an IBM System 36/38. At that time PC networking was still a 'toy' that nobody would use for serious accounting and the only other cheap alternative was Concurrent CP/M with its revolutionary overlapping sessions.

      5. In the ERP space, competitive opportunity generally only arises when there's a real/fake paradigm shift (e.g. DOS in the early 80's or client/server in the early 90's) or customers demand functionality that other packages don't offer. But I don't get the feeling that customers are demanding the 'seamless audit trail'. Which is a shame.

  • Colm O hAonghusa

    I think the following sums up the approach of the 'big boys' in this space.

    Several years ago I was subjected to a XBRL-related presentation by Oracle where the emphasis was on there being no need to consider a specialised database for storing instance documents etc: just shred the instance document into the existing relational database: easy peasy.

    Last November here: http://www.eurofiling.info/11th_workshop/11th_Wor…
    Mark Drake of Oracle gave a presentation ( not yet available) which ridiculed 1000% the prior exposition on shredding and promised instead Nivana, (or Valhalla, don't remember which) through the use of their new super-duper database which was still in development/testing. I found the unbridled arrogance of the presentation breathtaking, particularly given the audience contained a number of European banking supervisors, as the proposition left one in no doubt that if you did anything else you were a fool. The icing on the cake was the promise of a seamless migration from all shredded type databases to the new non-shredded!

    On the idea of a seamless (external) audit process, the audit firms will be slow to adopt this as it will attack their revenue stream for ticking and bashing.

    The idea of there being a seamless drill-down system for MIS is something I have attempted to achieve in a career devoted to delivering top class MIS systems, hence my deep interest in XBRL