We live in an information flooded world fueled by a constant state of “connected”. Businesses have tried for quite a long time to become “data driven”, so with the avalanche of data and ubiquitous connectivity you would think that supporting business decisions would be a routine approach at most companies. But, as all of you who just gasped loudly at the word “routine” know, we are far from being able to get the “right” data to the “right” person or team at the “right” time and with the “right” context. I’ve written about this subject before, and about decision systems so I’m not going to rehash that here, but instead I want to move on to a different part of this subject, how companies could use data and get relevant data to help in software buying decisions.
Software is critical to operating a modern business, which also means that the types and sources of that software have multiplied to an almost incomprehensible point. The way software is delivered is changing, and the transition from older to newer models adds to the complexity by creating hybrid solutions that need to be integrated. The idea of sourcing all business software from one or even only a few software vendors is an enigma from a different time, at least for large businesses. Suites, or more appropriately “clouds”, still exist of course, but they are hierarchal sets of nested clouds, wade up of many modules. For the mid and small business if is common to find the deployment of a broad set of functionality in a cloud suite, but even then there are many other specialty products needed to meet changing functional needs, and deeper vertical capabilities than horizontal suites can handle without complex customizations. I guess you could sum this paragraph up by saying simply that business technology is critical and often complex.
Dealing with this growing level of complexity for the business software landscape is a challenge for any size business, although the process does tend to vary by business size segment. Even with that variation of process the supporting data is fairly common across all the size segments. The basic flow of a selection process should have common elements, but the actual decision is related to organizational complexity and company governance and compliance guidelines. The steps should include, at a minimum:
- Awareness of business issue(s)
- Business needs analysis / research and identify potential solutions
- Develop functional and technical requirements
- Research solutions mapped to requirements
- Short list candidate solutions
- Evaluate short list
These “steps” are relatively generic by design, so as to present a broad enough framework to accommodate a variety of requirements and processes. There are more need/gap analysis activities post selection, but that gets into implementation so not really the subject of this post. A few other observations, first, this is not a linear process at all, so businesses will move forward and backward inside the steps as necessary. Second, you’ll see that I put research in twice, I’m not really sure if I should just call all 4-5 of the steps up to selection research, but calling it out in the way I have at least focuses on the activity that should be driving the whole set of activities. Third, don’t get caught up in the name of the step, nor in some artificial timeline, timing varies by business size, problem complexity, internal process, culture, governance controls, etc. In a small business the whole process could be handled by the “owner”, or in a mid-size or large business a team of many people, depending on the system being evaluated.
With that sort of simple process in mind, let’s look at data to support the process. The first part of the analysis can be initiated from almost anywhere, so the potential data sources there are very broad. It could be kicked off by a system replacement cycle, a change in a process, a need for greater efficiency, a breakdown in strategy or new competitive threats…the list is endless. The conversation around this is back to the decision system I mentioned above and wrote about several times over the past couple of years. Once you understand your problem or opportunity (all issues aren’t just created by a problem of course, it could be a new business opportunity that requires some system changes to support them), then you can start to use external data sources to identify ways to solve the issue(s).
When you’re looking for external data the list of criteria is fairly simple, it has to be from a trusted source, it has to be relevant to your current situation, and it has to be accessible at the right time. Trusted source is ultimately based on a personal decision but there are factors that can influence that trust. One factor that is pretty common to people in general is the idea of “like me”. In other words we look for advice and tend to trust opinions from people that are to some degree similar to us. Offline we have sought out people like us for ages, you join groups that have similar interests, are culturally similar, have similar experiences, etc., so it’s no surprise that our behavior online is modeled after this same concept. Trust is very high when the online source is from that ”like me” category. The PR/Marketing firm Edelman does an annual global survey and report on trust called the Edelman Trust Barometer. In that report they look at trusted sources of information and advice, including where people go for purchase support information. In general peer generated media are 2 of the 3 most trusted sources of information, search and social. When looking at categories of “spokespersons”, “people like me” was 3rd in 2016 at 63% (up from 57% in 2015), only behind academic expert 64% and technical expert 67%. What’s more telling though, is what is lower than people like me and how wide the gap is between the top 3 and the rest: financial industry analyst 53%, employees 52%, CEO 49%, NGO representative 48%, board of directors 44% and at the bottom, government official at 35%. That gap is very telling.
To support the data needs of researching solutions to an issue or opportunity, you can look to several sources that would meet the criteria of trusted, relevant and accessible. I talked about influence before, particularly in this post, so I’m not trying to rehash that topic. Instead I will look at some potential matches to the criteria for decision support data. Here are a few ideas:
- Public social networks: This is a good source of information if a bit variable depending on how connected you are to people that meet the criteria of course. Assuming that you tend to connect with people like you, then you’d assume that at least a part of your network would meet that criteria. Relevancy might be a little harder to find but again, I’d assume that it could be met, and of course accessibility isn’t likely an issue. I’d rate this as a highly useful source with a few caveats.
- Media (online and traditional): This is a bit more complicated since you’re searching and could have a little more difficulty getting to information that would meet all 3 criteria, but not impossible. Most of us have a few media sources that we already trust so it’s a matter of searching for content that is relevant to your specific set of questions. Accessibility might be an issue if the media was behind a paywall that you did not or chose not to support. I’d rate this one as medium useful as long as accessibility is dealt with.
- Industry organizations: A relatively good source of data, at least at the exploratory level of the research. This would depend on the level of relevancy and detail of the data, which varies by organization. Accessibility is restricted but if you’re planning on using it as a source I’d assume that you have or plan to solve the access issue. Since the nature of the organization is to be “like you”, at least at the industry level, this should be a medium to highly useful source.
- Industry analysts: For most industries there is a healthy set of analyst firms that provide service to that industry as well as the technology focused firms. Access depends on several factors but usually comes down to a paid relationship. The research will likely be helpful at some level but perhaps the more useful option is inquiry with an expert, and as the Trust Barometer shows, experts are very trusted. The information varies by firm, analyst, research methodology and experience so going into the inquiry with as much background on the firm and analyst is wise. You will probably not find complete alignment on the “like me” part of the data search, but if you have a trust relationship with the analyst the information would most likely be highly useful.
- Consultants: Many of the same caveats and observation apply to consultants as well. Both analysts and consultants should spend some time learning about your business and the issues before offering advice and information. With consultants generally the methodology allows more time for this phase of gaining understanding about your specific issues than with analysts. The usefulness and appropriateness varies by company and individual needs and always remember that there are a lot of consultants out there (and analysts too) so quality can vary so best to do a through bit of research yourself before engaging them. The outcomes can be highly useful, tempered with some common sense on your part.
- Vendor sponsored resources: There’s often a lot of valuable content and data available on vendor web sites. The issue may be one of trust however. Fairly consistently in surveys the vendor web site and sponsored content scores somewhat less trustworthy than independent resources. Even so, you can get useful information from the vendor and should check it if, and when you have learned enough to know possible solutions (in other words when you know the vendor is relevant to your situation). The variability of the resources will be great and you have to exercise an amount of critical thinking when you’re using them but they can be at least of medium usefulness.
- Vendor sales professionals: This is a difficult one since you inherently (at lest in a broad generalization) don’t trust sales reps. In a recent G2 Crowd survey (March 2016) 62% of the respondents reported that they only contact the vendor sales rep after they have made the purchase decision, which is certainly way after any data they would find could be used. I guess this is not very useful although it really depends on the relationship and the sales rep, not the generalization.
- Vendor reference companies: Vendor references can be vey helpful in the decision process if they are enough like you, and are trustworthy. It’s useful to understand the motivation behind taking the call. It can simply be that the solution is working and the person wants to share that, or in some cases there can be more to the arrangement…bet to be careful. In my experience the biggest issue is getting the “right” references and the “right” individuals to participate. If it meets the criteria then it can be highly useful, if not, then probably not very useful at all. Also remember that talking to 1 or 2 data sources is pretty limiting in the overall scheme of data collection.
- Peer review communities: Well I’m a little biased here since I work for one, but there is a lot of valuable data to be gleaned from peer reviews. From a references standpoint the crowdsourced nature of the reviews, which are in reality in-depth surveys, the information is much more likely to meet all 3 criteria. I do have some criteria for establishing trust of course, refer to this post for those. That post also does a good job of covering the types and use of information on leading review sites so I won’t go back through it here. I will just say that the better peer review sites do met all 3 criteria and are highly useful.
I didn’t spend much time on search engines, even though they are still rated very high on the Trust Barometer. That’s not a slight, certainly search will prove invaluable to your research, although I would caution that there is a lot of sponsored content there too, so use some discretion and common sense. I’m sure I’ve left out a bunch of other resources so if you think of one/some pls leave them in the comments.
(Cross-posted @ Michael Fauscette)