Marketing Week’s flawed IoT survey

marketing_week_infographicA few hours ago, Marketing Week published an article in their Trends section titled Smart Homes Lack Consumer Connection.  Although I’m an eager proponent of Internet of Things, I don’t find much insight or any actionable conclusions here for a number of reasons that I’ll explain below.  Do you find it insightful or helpful?  Does your answer change after you read this post?

What, no privacy concerns?

When it comes to people declining to install “smart” devices, the breakdown of their reasons as provided in the article is:

45% - Cost
44% - Unimportant
23% - Complexity
21% - Inappropriate data collection
18% - Intrusive
 3% - None of the above

Apparently it is possible to drastically reduce the ranking of privacy concerns by distinguishing between “too intrusive” versus “data being collected and used inappropriately.”   That’s quite a fine line to draw considering the lack of granularity in the other categories and instead of “Privacy – 39%” which would have trumped Complexity, we get two separate line items falling below everything else on the list except for “None of the Above”.  On the one hand it’s great that the study authors found something nuanced to look at.  On the other hand, gaaaaaaa!

What does “cost” mean here?

For example, a plethora of issues appear to be lumped into “cost.”  We all know that “cost” really means “cost versus benefit” and the article fails to distinguish whether people actually like the devices on offer and in their current form – i.e. see the devices as as highly beneficial.  Maybe respondents love the devices but lack the funds to buy them, in which case a plausible ROI demonstration is appropriate.  A good example of this is 40 watt equivalent LED bulbs that used to cost $30 ~ $50.  Now that they sell for < $10 they have gone mainstream.

That seems to be the direction the authors are going when discussing energy saving devices and use the phrase “save money” four times in the article.  On the other hand, “cost” may mean it isn’t worth paying the price for the devices on offer because the additional benefits derived simply aren’t compelling.  A good example of this was when there were no 100 watt equivalent LED bulbs or 3-way LED bulbs.  You had to pay a lot more money for something that wasn’t as functional as before.  Kinda like buying a “smart” bulb and then having to duct tape the wall switch to the On position and use your phone to control it, or having no control over a “smart” device when the Internet goes out.  Too bad the study authors didn’t see the need to find any nuance here.


Revolv Hub

This image embodies much of what’s wrong with IoT. Rather than replacing devices with functionally equivalent smart devices that provide enhancements, today’s IoT expects you to buy new types of devices, designs them as though you wish to feature them in your decor, and requires you to control everything over the phone.

Just as a raftload of sins are hidden under “cost” in the study, so too are they aggregated under “not considered important in my life.”  Does that mean “not considered important enough to find a place to put this new device on display so my friends will know how cool I am” (see the Revolv hub photo in the article) or “because I’m a Luddite,” or something in between?


Every single person who enables the buzzer on the washer and dryer has indicated their desire for those devices to notify them.  Everyone whose telephone is not set to mute, whose doorbell is operational, who use an alarm clock, who use a kitchen timer, have indicated a desire for notifications.  It is impossible to argue that notifications themselves are unimportant, so what is it about these notifications that is not compelling or relevant?  Perhaps it is because the notification destination is almost always the phone and that ambient notification devices are never used?  Of course, use of ambient notification systems would require integrations to a wider variety of devices and Industry seems to be well aware that Internet of Things is not about that.  No, the IoT is apparently about controlling, rather than enabling, all your device integrations.  That may be significant part of the problem but you’d never know it from reading this study which never considers whether the prevailing device architecture is part of the problem.  The article not only fails to provide any insight in this area, but it doesn’t seem to recognize that there’s any nuance to be found.


The other side of smart devices is actuation.  The primary time most of us wish for actuation is along the lines of “did I turn off the [insert name of device here] before I left the house?”  We’ve had device-issued notifications forever, even to some extent remotely, but we have not had a lot of “smart” actuation before.  For many people “not considered important in my life”  probably means exactly what you’d think and what the article suggests: we haven’t had this capability up to now and we don’t generally sit around wishing we did.

But “not considered important in my life” could also mean that the functionality of the devices on offer is perceived as laughable.  “You want me to replace a perfectly good wall switch with…my phone? BWAHAHAHAHAHA!”  This is the group into which I fall.  Admittedly this conclusion requires an informed and tech-savvy consumer.  However, targeting the portion of the market who do not understand the problem with this creates an incentive and business model based on keeping them clueless, and which also happens to facilitate the device-as-data-collection-portal paradigm.  Anyone but me have a problem with this approach?  Anyone else believe that devices should first act like the analog thing they replace and then provide enhancements as a secondary function?

It is also possible that “not considered important in my life” means “the device on offer doesn’t have the integrations that would make it compelling and traps me in a walled garden making it unlikely I’ll ever get the desired integrations.”    Call me crazy but when my deaf aunt comes to visit, I might actually want the doorbell, fire alarm, toaster, washer and dryer to talk to the house lighting so she can receive notifications just like everyone else in the home.  Anyone else believe that all devices should have open APIs so that prosumers and integrators can build compelling functionality with the mesh?  Or believe that a mesh of connectivity across all these unlike devices from different vendors needs to exist in order to realize the potential of IoT?  Maybe doing that would make IoT more relevant to the average consumer.  The study or authors, not sure which, or both, don’t seem to care how the “not important” category breaks out or whether the architecture is part of the reason people decline to buy.  Too bad.  We might have learned something by drilling into these issues.

Privacy – it’s in there

Marketing Week screen shotThe one area in which the authors found some nuance was privacy concerns.  It is unfortunate that the result of granularity in this category is to drastically understate the relevance of privacy in consumer minds as compared to the other categories.  The effect is apparent in the summary  that Marketing Week uses when referring to the article from elsewhere on the site: Consumers cite cost and lack of usefulness as barriers to adoption.  No, they didn’t.  If you combine both of the Privacy categories, there is a total of only 6 percentage points separating Cost (45%), Relevance (44%), and Privacy (39%).  Complexity (23%), which is the next closest category, comes in a distant 12 points below Privacy.  The concerns expressed seem to cluster around Cost, Relevance and Privacy as the barriers to adoption.  Odd that privacy would get dropped like that.

Perhaps when your audience is an industry driven by the collection and analysis of consumer data, to suggest that consumers have significant privacy concerns is taboo.  Or perhaps the researchers genuinely wanted to drill down in this area because it is important, created sub-categories for privacy, but that intention got lost in publication.  Hard to say what is going on and since the usefulness of the conclusions varies so widely depending on how you read the intent here, any credence we each give the study will tend to align with our own confirmation bias.  Anyone can interpret the results according to their own views and that, for me anyway, renders the results meaningless.

Does anyone other than me believe that devices should default to not sending data to the vendor and instead allow the device owner to optionally enable vendor access to the data based on receiving something of value in return?  That model would not only significantly improve consumer perceptions of data collection and intrusion, it would actually contribute to consumer confidence in IoT privacy.

Spin doctoring

Marketing Week Screen Shot

The article features an infographic, followed by this opening text.

I’m forced to make a lot of assumptions here because the study isn’t linked from the article and not accessible through Google search or anywhere else that I’ve found.  Since we do not have access to the study or information about its origins, we have to work with what’s in front of us.  Unfortunately, what’s in front of us doesn’t hold up well under close inspection.

Strangely, the first words in the article (at least those that aren’t a headline) are “The study, seen exclusively by Marketing Week, reveals…”  To which study are they referring, and what do they mean by “seen exclusively by”?

Are they trying to imply that someone independently and spontaneously funded this research without Marketing Week’s involvement and then gave Marketing Week exclusive access to it? The headline mentions “new research,” a non-specific phrase which could be plural or singular and suggests no connection exists between the reporter and the news being reported.  The rhetorical device of starting the article copy by back-referencing an unnamed but specific study from among all the available “new research”, and the passive construction using “seen exclusively by” combine to reinforce the suggestion that this is independent news reporting. So too do the references to “Source: Gekko” as the authors of the research.

If all that is true, then who commissioned the research?  And how did it end up as a Marketing Week exclusive and with their branding all over it?  Did Marketing Week vet the provenance of the study before publishing it?  Or did they in fact commission it themselves?  Why not just tell us the origins, scope and charter of the study or make it available, unless the intent is to deliberately put some spin on it?

To be fair, my suspicions of deliberate spin doctoring assume that the article was written by someone whose core competency is the use of English language in the art of persuasion, for example a marketing professional or experienced reporter in that field.  Someone like that doesn’t end up with a product like this by accident.  On the other hand, one could (some might say should) could give Marketing Week the benefit of the doubt and assume that the unusual rhetorical construction isn’t actually deliberate framing but rather a case of sloppy as hell writing and editing that managed to get past all the approvals required for a high-profile feature article.  Hey, it could happen.  Decide for yourself.  Got a different interpretation?  Let me know about it in the comments.

Personal conclusions

My issues with the methodology, the article’s interpretation of the results and the apparent framing lead me to conclude that there’s enough of an agenda showing through to distrust the whole thing.  I would have much preferred if the authors had drilled deeper into the broad spectrum of reasons consumers give for not buying today’s IoT devices.  There are very few devices on offer today that provide a combination of compelling functionality, an open API, operate when disconnected from the Internet, and integrate with anything.  Any study today would therefore be constrained by consumer perceptions of the crippled proprietary devices we have now as being representative of the possibilities of IoT, and thus such a study would be marginally useful at best.  But it would at least be more useful than the study presented.

Leave a Reply