Trend Report: Text Analytics in 2013
- 3/8/2013 5:48:04 AM
Sid, I would agree with you to a large extent, your point definitely makes sense and certainly many clients prove it right. On the other hand, marketplaces are emerging promoted by large players (this is the point we made in Seth's questions). Whether marketplaces are the answer that end users want (or single providers) is something we will see happen during the coming months and years. Personally, I'll be happy to watch it and discuss it together!
- by Sid Banerjee, Prospector
- 3/5/2013 4:18:17 PM
Beth - yes - I agree it may make some of those technologies ripe for acquisition.
The marketplace model works to a point, but I've found that vendors who do direct, or who license fuller products vs. depending on integrations into marketplaces provide higher value and command higher prices. For instance, Clarabridge (my company) sells the majority of its products direct, but we are also a Salesforce/Radian6 Insights Platform Partner (the marketplace mentioned by Antonio). The Radian6 customer who uses Clarabridge through the marketplace gets a very specific use of Clarabridge against the Radian6 Social Media content they're analyzing, but a direct Clarabridge Customer gets a full function multichannel text and sentiment analytics platform that's integrated into wide range of social, survey, community, forum, and even internal CRM platforms for multichannel customer analytics.
I'm a strong believer that if you want to maximize the value of your platform, you need to solve a solution value prop, and that means going direct, not depending on a marketplace that solves a specific set of problems for the marketplace vendor, but working directly to solve direct customer problems.
- 3/4/2013 5:11:55 PM
Beth, taking my answer with a bit of distance --since marketplaces are starting to consolidate now, and my area of expertise, computational linguistics, is far from marketplaces-- the marketplaces I've seen have significant diversity and number of participants. For example, we participate in the Salesforce Marketing Cloud marketplace, and it has around 20 providers; most providers do not overlap in functionalities but others do partially. Also, Temis is working on an initiative, in a way similar to a marketplace, more focused on semantics than text analytics; and my impression is that it will probably gather a higher number of providers and higher overlapping. It will also depend on whether the marketplace is fully open or requires invitation.
- 3/4/2013 9:30:16 AM
Sid, great insight. Thanks for sharing. You mention consolidation as the larger platform vendors look to obtain all the IP they need to be the "be everything" to their customers. Doesn't that make much of the more innovative technologies, algorithms, and products coming out of the open source and start-up environments ripe for acquisition?
- 3/4/2013 9:25:13 AM
Hi Antonio -- thanks for jumping in here. So let me ask you the same question I asked Seth, which is how much diversity do we see in today's marketplaces? Or, put another way, how far along in developing the coopetition/competition idea you've mentioned here?
- 3/4/2013 7:50:08 AM
Sid, good point and excellent description of the two scenarios.
My point: beyond the two types of situations you are describing (larger companies and smaller startups), there is a third variant --the one I was referring to-- where large companies create marketplaces to serve their clients using different text analytics providers. Examples of large companies promoting this type of marketplace are Salesforce or QlikView. My guess is that the main benefit they get is to provide their final clients with the best technology for their language and their needs (probably assuming it will be very complex for any vendor to have the best technology in all languages).
Happy to watch, and discuss, the market together for the coming months and years.
- by Sid Banerjee, Prospector
- 3/1/2013 3:35:49 PM
Great market overview, and good conversation here. I'll add my to cents on the market dynamic - I do think the text analytics market is moving in a few directions concurrently and one of the trends may favor marketplaces and market fragmentation, but the other trend favors larger company growth, and consolidation and less dependence on smaller vendors and marketplaces.
Companies like Clarabridge, SAS, Attensity, and others are growing (some faster than others) and are not looking to embed technologies from third parties so much as become the master vendor for all things text analtyics (data sourcing, data processing/classification/sentiment/scoring/quality refinement/analytics) and are seeking to create value by developing and owning all the core IP themselves. They are less likely to want to go to a marketplace as to build and own the key components they believe will create the highest quality, most robust platform for text analytics. Having done that, they beleive that they can then develop the most robust use cases of the platform, either for themselves, or by licensing those platforms to partners. But these companies aren't generally using marketplaces, they're bigger companies, and selling through more traditional direct selling and partner selling models.
On the other hand, there are many rapidly developing technologies, algorithms, and products that are being developed in the open source community, by smaller startups, and that are spinning out of services companies. These technologies, which are more "widgets" than fullly developed platforms, are finding users, partners, and integrators (as well as non-text analytics licensors) through a marketplace medium.
If you're a marketing, customer care, or customer experience buyer you're more likely to buy one of the platforms from the bigger vendors. If you're a technology developer working for a software company, a SaaS vendor looking to embed new algorithms into your cloud offering or a systems integrator looking to build a custom text mining app - you're more likely to go to a marketplace model so as to have the maximum number of alternatives and competitive options, as others have suggested.
There's room for both modes of finding solutions, depending on what you are looking for.
- 3/1/2013 10:23:50 AM
Seth, thanks! Happy to contribute. Beth, thanks for your comments.
In a nutshell, these are the reasons why we think marketplaces are a relevant development for text analytics as a base functionality for other end-user services, like Social CRM, EFM, VOC...
1. An environment like a marketplace fosters coopetition (competition/cooperation) which will most likely result in better accuracy, a common problem nowadays in text analytics, and as a result in higher end-user adoption
2. Decoupling the two tasks (text analysis, and insight extraction & display) will result in higher specialization in Social CRM, EFM, VOC or other services, hence, more focus on each provider's specific tasks and normally better service for clients (i.e. a Social CRM company like Salesforce can concentrate on insights extracted from text analysis rather than on text analysis itself, which a highly specialized task)
3. And marketplaces also open the door to multilinguality, since different providers will provide different languages (no text analytics provider has ALL languages), removing an adoption obstacle that is relevant for some global companies today