The pros and cons of using end to end software in market research

Market research software products have expanded over the last 20 years to cover online surveys, tabulations (crosstabs), reporting, online dashboards, CATI, CAPI, paper surveys, online communities, social listening, text analytics, automated coding, text processing and analysis and more. As a result of this, there are software packages or platforms that cover one or more of these different tasks. So, should you buy as few systems as possible or buy the most suitable product for each of your needs? The decisions are getting harder as time passes, so let’s look at the pros and cons of using end to end software in market research.

Scope of this blog article

This article is focusing on the more traditional aspects of market research covering from data collection to analysis and reporting. Does it make sense to try to find a system that does everything – or almost everything – and minimise the number of products that you use? I will argue that it depends on what you are doing, which might sound as though I am evading an answer. However, I will try to look at each of the factors that should influence your decision. So, let’s get started.

1. Learning and expertise

One of the obvious advantages of using an end-to-end system is that everything is in one system. This advantage makes learning easier and, perhaps more importantly, means that a group of people can become experts or highly competent in using the software. It is likely to mean that a wider group of people know ‘how things work’. With several products covering each part of the process, it is more likely that smaller groups or an individual are responsible and have that knowledge. Having a spread of knowledge makes absences, staff leaving or peaks of work much easier to handle; it is an important consideration.

2. Data compatibility

Another advantage of an all-in-one system is that data compatibility is not going to be a problem. You can expect, one would hope, that moving data from the data collection to the analysis and reporting phase is seamless and requires no work on your part. If you use more than one system, you will need to look at how easy it is to move data from one system to another. I have seen data processing teams spend hours or days recoding data to meet the requirements of a second system. This potential problem needs to be checked before implementation as data processing staff can have a mentality of conquering challenges rather than telling management that it is difficult.

2a) Common data compatibility problems

Some of the most common problems arise from things that seem fairly routine in market research. The top things are:

• Handling of multiple response questions – there are three or four ways of storing data from such questions and some software can only handle one of these ways. Transferring data to another system may be difficult or, even, impossible.
• Handling of numeric values – some software may have problems with decimal places or rounding.
• Handling of blank/zero – some software will interpret blank as zero, others will treat it as a missing value. This issue might sound trivial, but it is problems like this that can take time to resolve or cause errors.
• Grids/rating scales – different software systems store this type of data differently.

2b) Data structures

Repetitive data structures within data are handled quite differently by different software packages. By this, I am referring to hierarchical data or blocks of data that are repeated. For example, you may have a survey with hierarchical data where you interview one doctor and, say, up to five of his patients. Or, you may have repeated blocks of data, where you ask a series of questions for each brand that someone uses or where you collect data for each time slot over several days. How this data is stored may vary significantly from software package to software package – or, may not be possible or very awkward to handle.

3. Data interchange formats

There are two commonly used data interchange formats where texts and data can be transferred smoothly from one system to another. The first is Triple-S which is managed by the ASC, an independent organisation. This interchange format is generally a good way to transfer data, although it does have a few limitations and I have experienced poor quality exports from some software systems which cannot be read back without manual editing. A second option is to use the SPSS format. This format is the proprietary file structure of a commercial software supplier. This format works well in many cases. However, checking these ways of linking two software systems is important to check before purchasing, in my opinion.

4. Productivity

One of the most underestimated things in a software purchase decision, I believe, is productivity. Yes, it’s hard to calculate, but as most other business decisions look at revenue and costs, there is no excuse not to try and measure productivity. This is where having a specialist tool can make a difference. If a specialist tool can, say, halve the staff time needed, it can soon pay for itself where there is sufficient usage, and the price tag is not sky-high. I have seen people struggle with software that can do something, but where it takes an unreasonable amount of time. If that’s the case, looking for a separate software package makes sense.

5. Skills match

The level of usage of a particular aspect of the process and the skills available should also have a significant effect on what software you buy. If your company spends a lot of time preparing online reports, it may be that you need a specialist tool unless the all-in-one system does everything you need. Often more complex software needs more skilled staff, so questioning the skills you have available and whether training can make staff more able to use smarter tools is important.

6. Complexities

End-to-end software can suffer in two often connected ways. It may have a weak link – for example, its tools for data collection are first class, but its tools for tabulations are highly limited. Secondly, its origins may have been one or two parts of the process, which means that the other parts are inadequate in most cases. I know many of the free/low-cost online data collection systems have very limited capabilities when it comes to tabulations and reporting. Yes, they can, perhaps, generate a chart, but whether that chart is of any use to your processes may be doubtful. If your work is too complex for an all-in-one package, it doesn’t mean you shouldn’t buy it, but you need to be ready to purchase something more specialist – perhaps, if and when your need increases.

7. Upgrading

Thinking ahead is not easy with software. However, I think it helps to think ahead and try to imagine what you might need in the future. It can help you consider what upgrades you might need in the future and how that might fit into what your end-to-end system or product mix work.

8. Cost

Cost is an issue in any organisation. For those who need to seek budgets, it can be difficult to get a budget to buy an additional product when the one currently used ‘does more or less what is needed’. The productivity argument has to be the key to this unless, of course, not having the right software threatens to fail to deliver clients what they want.