Long range trends in small business accounting software: GL dialtone


Nearly 100% of small businesses today are running standalone or LAN-based accounting software which is

We are at the stage where mechanical sewing machines first appeared. They became wildly popular for a few decades, until better equipment was invented for making clothes.

Small business now wants the following, rather obvious, communication abilities:

The above needs of small business are blocked by the lack of standards, more than anything else. If all of the shrinkwrap accounting software vendors awoke tomorrow morning and released electronic billing modules, which used email or HTTP to transmit and receive invoices, millions of businesses would begin using those functions within a year or two. Frankly, I don't understand why progress has been so slow.

In addition to conducting business over internet, the small business market also needs functionality which is not easy to retrofit into LAN-based packages:

When you consider the commercial needs in the small business market, they point to a future where many accounting systems will take the form of a web server, fed by an accounting database. The database and functionality of the accounting system will not change much --but the technology for delivering the user interface to the screen will become the web browser rather than the operating system. A high quality user interface with no less richness or speed than today's software can be supported by either Netscape or Explorer. On a LAN the perceived response will not be different than today's fine-grained database connections.

The redesign of accounting software to HTTP transport will be driven by other forces besides the need to conduct selling on internet. Remote access will be an immediate objective of any web accounting architecture. Integration with other software would also be easier-- HTTP is widely understood.  Software developers could design the GL to operate  thru native HTML pages, or enhanced with Javascript, VBScript, DHTML, or accessing ActiveX objects with MS Internet Explorer. 

The design of any new web-server based accounting system with mature accounting features, particularly in complex areas like inventory and project or job cost may take 2-3 years. The real question is, from what segment of the software market will such general ledgers emerge?

I believe it will prove easier for developers of E-commerce software and websites such as storefronts to bolt on accounting functions, than for LAN-based accounting systems to bolt on the internet. For example, Great Plains, AccPac, and other higher-end accounting systems have Web storefront modules well integrated into their family of modules but the entire architecture including remote access, etc. is extremely complex and expensive. These  high-end accounting systems technically may be able to support some aspects of internet commerce, but the system as a whole is not a practical, well-balanced solution for it.

These GLs might be grafted onto more generalized E-commerce software or websites, forming a total solution that could meet all of the needs of multiple small businesses, listed above. However, the systems skills, software licensing, and complexity and ongoing labor costs would price this service beyond most small businesses.  Think of how difficult and expensive it is to move from 5 to 10 or 30 or 100 users, with any of these accounting software and you realize the truth:  client/server is little better than file/server. You pay 10 times more money to add maybe 4x more users.

I see little evidence that enterprise accounting vendors Great Plains or SAP will be picked up by significant E-commerce hosts, as inventory or accounting components. These ERP vendors have little presence in the small business market at current pricing, and they actually do not have thin client technology let alone XML technology or HTTP technology. They run remotely only by the 3rd party "PCAnywhere" technology of Citrix Winframe or Windows terminals, which passes the keyboard, mouse and screen data across the network. They have excessive program logic on application modules, to fit a Web host model. The keel for this battleship architecture was laid in the early 1990s, and has a large installed base which cannot easily change course.

ISPs and web hosting companies are currently facing difficult challenges of inventory management and selling over internet. Those who succeed will find it relatively easy to develop the remaining portions of accounting functionality needed to provide small business with a complete accounting database.  For most businesses, 99% of the accounting transactions are in the sales journal (AR) and purchases (AP), and the   associated inventory, collections and payments entries.   Web storefronts and i-commerce providers who ever rise to the challenge of managing these transactions   will certainly wake up, and provide an interface for accountants to make the last 1% of the transactions: Journal Entries.

Web-based GL software is beginning to appear in at least 9 sectors:

1. ERP software appearing on ASPs or subscription hosts over internet. See
   list at http://www.gldialtone.com/EnterpriseGLdialtone.htm
2. Similar lower-priced online GLs aiming at small businesses or niches
   (a few examples are at http://www.gldialtone.com/GLdialtone.htm)
3. E-commerce hosts (ISPs) extending their functionality to support 
   general ledger tables.  Web hosts capable of online selling,
   maintaining inventory, and security and payments technology, adding 
   accounting features and ultimately, General Journal and Financials.
4. Online Time-and-Expense vendors are relatively mature, adding general
   ledger interfaces or components to link with corporate General Ledgers.
5. Traditional LAN-based accounting vendors bolting on "the internet"
   to their existing, robust accounting systems.  These systems are fundamentally
   database-connected and fat-client software.  But they can be accessed over
   the internet by VPNs and Citrix Winframe, and they can be installed alonside 
   web servers for internet commerce. http://www.gldialtone.com/bolton.htm
6. Web development tool vendors extending their products to include decently 
   integrated inventory databases.  This provides a platform for features such as
   purchasing and vendor tables, and finally, General Journal and Financials.
7. Free/open source software projects http://www.gldialtone.com/sellGPLservice.htm
8. In-house software.  New, intranet-based development in corporate environments. 
   webservers.  Vertically integrated applications including General Journal are
   appearing in architectures such as ASP/ MSIE4 or MSIE5, or Java beans.
9. General ledgers and billing systems developed specifically for ISPs, to
   support their particular billing needs integrating network accounts, security
   email correspondence, electronic billing, etc. See the ISP-Billing website
   and Shawn Hogan's product comparisons


Thus, I predict the following model will emerge in the short term:

1. Ambitious E-Commerce hosts will undertake to provide a database back end for accounting, with SSL security.

These hosts will dish up responses to queries from accounting users, in plain HTML or XML forms. These will including familiar web input boxes, and screens full of transactions in tabular format, for the user to read. The pages will be sent in XML, the emerging tag language that can be rendered by a browser with XSL, but can also, simultaneously be machine read by any XML client software.

The web host will be a repository for data, and will serve up its rows, sorted and filtered and joined according to whatever the HTTP client requested. The RDBMS back end on the web site will provide additional value unavailable on client sites in the near term:

Early leaders in the Web ledger market will get a foothold within the shelter of emerging e-commerce standards, for example Biztalk, CBL, cXML, OBI, XML/EDI.   Accountants are not the prime mover behind the powerful e-commerce thrust, but rather, beneficiaries.  Accounting system developers will observe developments on XML DTDs and schema.  These  standards and implementation details will provide the necessary framework to advance an open, generic standard. I have been calling this GL dialtone, for lack of a better term.

2. Ambitious programmers will build client software which would read the standard XML pages sent by the E-Commerce host with any XML parser, in any language they like. Examples of such client software could be

  1. simple custom applications, for certain clients or industries
  2. Active-X components or other wrappers for XML, making the technlogy useful and accessible to Delphi, VB and Foxpro programmers, Solomon VARS, Goldmine VARs,
    telephony VARS, and countless vertical market developers.
  3. addon components or connectors, connecting existing software such as Quickbooks or Peachtree to the GL Dialtone provided by E-commerce hosts. 
  4. Translation modules or interfaces, translating CSV to XML, etc.
    In other words, simple, freestanding widget which you drop in ASCII, and it outputs
    XML accounting data files or streams in formats like XBRL.

3. Ambitious service firms such as CPAs and payroll and bookkeeping providers will get into the technology, identifying programs or elements they need, and begin servicing their clients with them, instead of Quickbooks, Peachtree, etc.

The immediate benefit for these "financial outsourcing providers" will be the ability to see and maintain the client's data at any time during the year, and for the first time, obtain views of transactions which have been added or edited since the last time they logged in. This is essential to an efficient review process for controllers, as well as outsourcers. CPAs and financial analysts will also achieve a single interface between the accounting data of their clients on the GL dialtone server, and the core software of the service provider such as income tax and financial reporting software. For these reasons, service providers will be an active influence in moving clients to Web GL's.

For their part, the clients will find a whole new universe of competitive, remote CPAs and bookkeeping services associated with the Web hosts. They will be able to switch service providers and take their entire reporting and data context with them, in a matter of seconds, anyplace in the world.

In summary:

The major enabling factors driving a transition to GL dialtone on Web hosts are:

The major financial urgency driving small businesses into these new accounting systems is

Significant factors exist, slowing the above architecture, but they are not fatal. These include the poor usability of web input forms and other simple HTML user interfaces, if applied to accounting users' desktops.

This is changing rapidly with innovations such as DHTML and Javascript. There is no intrinisic reason that a Web Ledger cannot support a user interface of equal quality and polish as Quickbooks, but it will take time to develop that quality of user interface.

It is not clear that the need for highly refined user interfaces on accounting software will continue to dominate software designs. If sales are received electronically from customers, and processed by the software, rather than by users, there is less need for slick "invoicing modules". Similarly, nobody will need bank reconcilation or even AP/AR reconciliation tools, in a world where transactions are created equally on the systems of both trading partners

Web GLs will also gain a free ride from the reporting interfaces of best-of-breed report writers like Crystal Reports and F9.  These could easily provide better reporting by Web GLs than what's on existing low-end accounting software.

Owners' concerns for privacy and security will give pause but will be overcome in almost every case, where tangible advantages exist. SSL is reasonably secure. Offer a small business the choice between 100% safety, or $1,000 plus 99.9% safety and they will take the $1,000 savings every time.

Another concern that I believe, will be overcome and forgotten, is the owners' cultural bias to locating his accounting records on his own premises, in a machine he/she can touch and feel. The existence of a competitive market in standards-based GL dialtone will reduce fears of poor service, price gouging, etc. that existed in earlier "service bureau" architectures. The exchange of invoices and other business documents will be intrinsically based on standards.

Psychological resistance would also be overcome in either of two ways:

An accounting architecture centered around E-commerce web hosts and their databases, provides a division of tasks between the hosts, VARs and independent programmers, and independent bookkeepers and accountants, that fit well into existing skill sets, commercial realities, and customer bases.

Todd Boyle CPA, Kirkland WA - http://www.gldialtone.com