InfoSENTRY  logo
Election TechREPORTS
InfoSENTRY® Services   www.infosentry.com   info@infosentry.com
Issue 13  
November 2002                          Privacy Policy

  FEATURE 1:  COTS, MOTS, and NOTS
  FEATURE 2:  
Rating Your Website's Accessibility

COTS, MOTS, and NOTS

There is a phrase election officials will see frequently in the coming months: Commercial Off-The-Shelf Software (COTS). You will see it in contracts given to you by software and system vendors for statewide voter registration software. COTS means that you will buy the software license, install the software, and it will meet your needs. That's it. Nothing to it.

At its core, purchasing a COTS system assumes that the way you will do your elections business (the workflow, the basic operations, the report generation, the poll book preparation, the database maintenance, and a raft of other functions) will be the same as the way other states and counties do their elections business. Kansas State Capitol Purchasing a COTS system, while assuming that it will not need much if any modification, also assumes that your laws, your electronic interface with the DMV, and your electronic interfaces with other NVRA agencies are fundamentally the same as the other jurisdictions in which it was installed. COTS means there will be very little or no modification of the software. You will modify your business to conform largely to the way the system structures the business flow.

You will see the phrase COTS, or a reasonably well-disguised version of the phrase, for two main reasons. First, it is a way the vendor has of formalizing in the contract an expectation that the software system will not be modified. Second, it usually allows the vendor to "book" the entire value of the sale for accounting purposes at the time the contract is signed. In other words, the phrase COTS provides almost full benefit to the vendor and no benefit to you, the customer. 

In our work with large system implementations and with the attorneys assigned to work on the projects, we recommend strongly against allowing the COTS (or a disguised equivalent phrase) from being in the final contract. The simple reason is that the basic assumption that the software system will not be modified from county to county or state to state is not valid. And you have very little to gain from helping the vendor put the entire sale on the books before your get your system modified, installed, tested, and in production.

Instead of COTS, the system is much more likely to be MOTS--Modified Off-The-Shelf Software. The modifications might be small. However, even small modifications to the underlying database structure might require hundreds of hours of customized programming. Larger modifications to database structure, system workflow, document imaging capabilities, screen design, and report structures might require thousands of hours of programming. 

So, the COTS purchase turns into a significant MOTS implementation project. Your original budget estimates are blown sky-high unless your have protected against this kind of escalation in your contract. You need to know upfront if yours is going to be a MOTS implementation and if the vendor has already priced the modifications in the contract cost.

When the project reaches this stage, requiring tremendous customization, it becomes a NOTS implementation--Not Off-The-Shelf Software. At this point, your project starts to look like the "normal" software lifecycle development project depicted in the following figure. This figure shows a very rough approximation of the software development path for a new statewide voter registration system on the top of the calendar. On the bottom is the hardware and network schedule.

Of course, the two schedules do not run as neatly parallel in real life. The individual stages will vary from what is displayed. Some stages will last longer and some will be slightly shorter. Statewide voter registration systems are too big and complex to flow that smoothly. And there are many other activities going on in the project at the same time: documentation, security and business continuity planning, quality assurance, training, organization change management and overall transition management--to name a few of the efforts we mentioned in last month's newsletter.

The impetus behind accepting the so-called COTS approach is to "cut to the chase," buying the hardware and network components, taking delivery of the software "out of the box," and telling the end users when to expect installation. It is also an almost guaranteed path to project failure. Time will run out much sooner than expected. The money will be gone. The system will not be near to completion. Enter the attorneys.

So, COTS often turns to NOTS by way of MOTS--and the lower "hardware/network" timeline in the figure above can often stretch to surpass the upper "build" timeline. There really is no shortcut by buying a package with the expectation of only minor modifications. To do so is to assume that your needs, requirements, procedures, and laws are the same as those across the border in another county or another state. 

Negotiating the COTS phrase and expectation out of a contract--and realizing the complexity of the task ahead--will be difficult. There is no need to doom the project from the very beginning by setting unrealistic, inaccurate expectations with a single acronym like "COTS" in the contract. All statewide voter registration system implementations from a purchased "software package" will involve some degree of customization and modification. In most instances, they will involve significant amounts of software changes and programming changes. Setting the expectations at a realistic level at the project's outset  might be critical to the success of your statewide voter registration system implementation. 

InfoSENTRY consults with clients and major law firms nationally to manage information technology implementation procurements and projects to minimize the kinds of pitfalls described in this article.

Return to top of page graphic
Return to Top of Page

Rating Your Website's Accessibility

Controversy swirls around various requirements for websites to be accessible. Federal regulations. Suspension of federal regulations. Court rulings.

Of course, this issue is part of a larger accessibility question with which the elections community has been dealing for some time. Campaign Finance Online Graphic The issue of website accessibility is likely to grow in elections offices as the new Federal election reform legislation extends certain Federal regulations to statewide voter registration databases. 

It does not take much imagination to hear the following argument: 

Text Bullet Federal legislation mandates that State Chief Election offices are the "custodians" of the "official" electronic records of every registered voter in the State. 

Text Bullet If the State uses any federal funds to develop a web portal to enter, store, or retrieve these official voter registration records, that portal must be accessible under applicable Federal accessibility statutes and regulations. 

Has it happened yet? Probably not. Do we say it is going to happen for sure? No. Could it happen? Yes. Is it a reasonable expectation to take accessibility design standards into account any way? Yes, again.

There is assistance on the market to help you improve your website's accessibility. While InfoSENTRY does not sell, distribute, or endorse any products, we recently tried a free demo of an "accessibility evaluator" (called Alice the Accessibility Advocate) found at www.ssbtechnologies.com

We submitted our website to "Alice" for a free review of its accessibility. The summary results came back quickly and cleanly. We were pleased to find that our website scored well on Alice's rating scale--but not high enough in our opinion. 

So, we used some of the data and analysis provided by "Alice" to modify our pages right away to make them more accessible. So, why wait to hear from an advocacy group (or Federal oversight agency) that your voter registration or campaign finance web site is not accessible by "up to 54 million Americans with disabilities" as the website says. Your office might benefit from an ounce of prevention at the design stage. It could save you a substantial amount of money down the road.

Return to top of page graphic
Return to Top of Page

Please click here to visit our main election systems consulting page. It has a table of contents for previous newsletter issues.

Please click here to visit our main project management consulting page. It has examples of elections systems projects with which we have assisted clients throughout the country.

Please click here to read a copy of our white paper on User Acceptance Testing

Please click here to read a copy of our white paper on After Action Reviews as a formal--but inexpensive and quick--way to capture lessons learned in information system projects.

Please visit our main information technology consulting page and our information technology security and recovery page. They contain brief descriptions of some of our previous consulting engagements, including those for election jurisdictions.

Please contact et@infosentry.com if you would like to get a PDF version or a laser printed copy of this newsletter for distribution in your election office.

InfoSENTRY Services, Inc.
2 Hannover Square, Suite 1740 Raleigh, NC 27601
P.O. Box 28048, Raleigh, NC 27611
Phone: 919.838.8570
Glenn Newkirk's e-mail:
glenn_newkirk@infosentry.com

Copyright 2002, InfoSENTRY® Services, Inc. All fights reserved. Reproduction and dissemination without the express written permission of InfoSENTRY® Services, Inc. is strictly prohibited. InfoSENTRY Services, Inc. publishes Election TechREPORTS monthly, focusing on technology trends and issues in election offices.  From time to time, Election TechReports might mention the name of vendors' hardware or software products. However, InfoSENTRY® Services is completely independent from hardware and software vendors. Mentions of vendors' hardware and software products in no way constitutes an endorsement or indication of worthiness for those vendors or products.