Article #17: Pollux, Bellview, Quancept, Snap and CI3


As CATI users spread, Tim Macer surveys the pros and cons of the rival systems now on the market

You pays your money - so choose with care


Of all the IT decisions that a market research company can make, choosing a CATI system is likely to be the most costly. There are many to choose from now, each with subtly different strengths and limitations. Other people's recommendations are not necessarily the best guide, since what is cost effective, or appropriate for them may be expensive and disastrous for you.

There are some basic differences in the way each system works. The original ones, including Bellview and Quancept, use a central mini-computer (not a PC) to drive "dumb" terminals. The kit can be pricey, as can software licenses, but the cost of adding extra stations is low and there is virtually no upper limit. These systems are technically quite demanding too, so the learning curve tends to be even steeper than the price curve. Later systems, such as Surveycraft and CI3 exploit the "client server" architecture of a PC network. Processing tasks are shared around the network. Being PC based, and coming later, they tend to be simpler to set up and use. For demanding users, this simplicity can become a lack of sophistication. Cost follows more of a straight line, as you add more stations. Where it crosses the central system's curve depends on many factors, but is probably between 40 and 70 stations. Beyond that, central systems are hard to beat on cost and performance.

CI3, from Sawtooth Technologies is a true networked system. Joe Curry, Sawtooth's president readily admits that CI3's market is small to medium-sized sites. "With a large number of stations, central is cheaper". But he pointed out that most central systems are "legacy" systems which have been ported to run on other platforms, and rarely have the genuine ease of use or felxibility of PC products like CI3. This was backed up by Ron Clark, who runs operations at IHR Research in Las Vegas. "CI3 has worked out tremendously for us" They have integrated it with their own office automation applications and systems they have developed themselves. Such development is much easier in a PC environment. Mr Clark dismissed any arguments over the unreliability of networks, saying they average one failure every six months. "The only significant disadvantage is the time delay you can get when the network is overloaded. But an intriguing advantage is it's a very inexpensive way to get into CATI. Their pricing makes it hard not to choose CI3".

Survey Methods Group, in San Francisco, recently made the switch from CI3 to Pulse Train's Bellview. Jeff Henne, director at SMG said "at the time we got CI3 it was absolutely the right choice. For 20 stations it works just fine, but some activities took a long time. In Bellview it is done in seconds while in CI3 it could take minutes. We went to Bellview because it could handle big sample databases and a multiplicity of surveys without these kinds of delays." His overall verdict, having made the switch is that Bellview is a lot more stable and they can do more complex surveys, but he added, "there is a big price difference between CI3 and Bellview, and I still wonder if CI3 wasn't faster for programming."

Roger Calverly, Director at Research Services in London is very happy with their decision to switch to the Quancept interviewing system, from Quantime, a few years ago. He said "We chose a centralised system because that was pretty much what was on offer when we first looked". RSL used Bellview before moving to Quancept a few years ago. I asked Mr Calverly why they chose to stick with a centralised system. "We moved to Quancept to match the software we use for analysis [Quantum]. Also Quancept had a big price advantage with their UNIX boxes. Using a network seemed like simply introducing complications. Though a disadvantage with a central system is that it can go down [when everything stops] and it can get clogged with work.

"Our whole experience of networks is that the more you let people mess around with PCs, the more problems they cause for your systems department. UNIX is very much more robust that MS-DOS. I think you get more productivity with it overall." Mr Calverly's overall assessment: "Quancept suits us very well. It does everything we want."

Define Research and Marketing International were new to CATI interviewing when they chose Pollux, a PC system developed in France and distributed in the UK by Merlinco. Mike Lorimer, Define's IT director explained "We started by looking for something that would run on Macs, so that it would integrate with all the creative work we do." Although they failed in their search, they are very happy with the solution they adopted. "We needed a dedicated system and I wanted to minimise maintenance for myself. We looked at Bellview, but it seemed complicated. We wanted something we could use internally. Quancept seemed good, but with the survey management, you have to write it and compile it yourself. In Pollux all the nuts and bolts are already in there. We also chose it for its cost and the training time. Having learned it, I found it to be very teachable". Interestingly, although Pollux is a true PC system, Define operate it as a centralised and not a networked system. Mr Lorimer's overall experience with Pollux was "very positive".

MSS Marketing Research in Bristol rather surprisingly selected Snap from Mercator to run their 20-station CATI operation. Snap is not an obvious CATI contender, as it is primarily a data entry and analysis tool. It operates independently on each PC and does provide automated sample management or quota control. Martin Olver, associate director at MSS, told me "We've basically grown up with Snap. All our secretarial staff set up questionnaires in Snap so it was not much of a leap to do CATI. The latest version is very good at converting your thoughts into a proper questionnaire. Snap does all the hard work for you. We've found it a low risk means of offering a CATI service with most of the benefits of larger systems". His verdict "Allowing for the limitations of not putting complicated surveys on it, I'd rate it 8 or 9 out of 10".

It pays to look around when shopping for CATI systems. There are startling differences in price and functionality. Not everyone is as candid about the strengths and limitations of their software as Joe Curry. But then CI3 has sold over 500 systems, including 200 in Europe - more than most. And the developments continue. Quancept and Pollux are working on 32 bit Windows versions; the excellent In2Quest CAPI system is sheduled for a CATI release, and major enhancements to CI3 are due soon. Choosing the right system just got harder.


Published in Research, the magazine of the Market Research Society, August 1997, Issue number 375.

© Copyright Market Research Society/Tim Macer 1997. All rights reserved. Reproduced with permission.

Top | Index