This really isn't up to iRobot Corporation or Colin if the company gets acquired. They are publicly traded, hostile take overs are possible. The promise is meaningless.
Wizards of the Coast made a public announcement back in the 90s that they would not reprint cards on the reserved list in Magic: The Gathering. Even though they're now owned by Hasbro, they're still afraid of a lawsuit from collectors and trading shops if they do reprints or functional reprints (because that would drop the value of existing collections, causing material damages).
They toyed with non-playable collector editions for a few years, and then even that was shut down.
Right now, the legal system does not always afford breaches and violations of privacy the level of gravitas that they deserve.
The challenge in this case is assessing the dollar value of a new piece of personal info (room geometry). Perhaps there should be a small floor value (priced in $/byte) for any personal data when assessing damages - it would greatly discourage collecting novel data about users "for future use" or "just because we can".
Edit: "anonymized" data can be discounted based on an expert assessment of how easy it is to de-anonymize.
- Publicly stated that they regret the reserved list, it's a mistake, and its existence hurts the game.
- Implemented a policy that employees are not allowed to publicly discuss the reserved list.
- Released tournament-legal cards from the reserved list in a manner 100% compliant with the terms of the reserved list.
- In response to complaints, strengthen the reserved list policy so such releases are no longer allowed.
This makes no sense to me as a criticism, because (a) the legal situation would be entirely identical if Magic: The Gathering were produced by a single individual, and (b) Wizards is refraining from printing more cards in response to pressure from the collectors.
Maybe I can restate my comment as "This promise is meaningless, but there are meaningful promises they can make that they chose not to, and that is meaningful."
[1] I don't think that is a word but hopefully what I hope it would mean if it were a word is clear :-)
As a tangent, there's also mechanisms you could use besides something in the bylaws to prevent it -- I don't think you actually need anything but restructuring the contract. (Though, obviously, bylaws or articles would carry more weight.)
[1] I like that phrase!
There are two choices the analysis; 'do nothing' which leaves the business as it is, collecting sensor data from the operation of the robots into aggregated data sets, and 'destruct on sale' which is a promise to delete any and all sensor data collected so that the new owner can not use it.
I am further assuming that at the time of sale the business has two components of value, its 'operational' value which is the annual recurring revenue of its subscribers/users less the cost of managing the revenue, and 'residual' value which consists of company assets such as the physical plant (offices), algorithms or patents, real estate holdings, and data sets.
In the 'do nothing' case we assume that subscription rates ($SUBS) stay the same (so revenue is unchanged) and that the data set accumulated has some value (say $DS)). Using $SUBS as the revenue model, $ASSETS for the stuff that doesn't change in either choice (office, patents) the value generated by the company would be something like:
years * $SUBS + $ASSETS + $DS.
years * ($SUBS + $Delta) + $ASSETS.
So my analysis would say that the 'better' choice is the one that generates more value over time for the company. And given the structure above that depends entirely on one thing, is the value of ($DELTA * years) greater or less than the value of $DS.
Next I'm going to stipulate that the value of $DS has a 'years' component as well since the data set keeps customer data regardless if they are a current customer or not. So the more customers you've had over the years the more value $DS. And using that stipulation I'm going to simplify the analysis to just the comparison of $DELTA and $DS.
Then I'm going to ask can we characterize the size of $DELTA? And I will argue, perhaps unfairly, that $DELTA would consist entirely of people who 'care about privacy.' Then I'm going to look at all of the experiments in tech where the differentiating factor of the product was 'privacy' or 'less tracking' and compare their addressable customer market to the the overall market. (I actually got to 'live' this advocating the privacy features of the Blekko search engine to people)
Then I'm going to argue that while there are many people who will talk about their concerns around privacy and security the number that will change their behavior relative to that concern if astonishingly small. It shouldn't be but there are examples all around of how it is. And based on that argument I'm going to argue that $DELTA is very small, 1 or 2% of the subscriber base at best.
Then I am going to talk about the challenges of developing a dataset like $DS. How difficult it would be for anyone to create such a data set without a way of explicitly running a sensor suite in thousands of homes. And given how hard it is to compete, that someone with sole possession of that data would have a competitive advantage over others who might seek to use it, or might have a salable product based on that data that nobody else could offer. (joke idea: use it to validate square footage numbers in Redfin listings). I would argue that this data set which would be unique in the world would be very valuable, and assert that $DS was >> $DELTA and so the right choice for the company was to keep it.
And in that analysis, if the most value would be maintained by keeping it, then changing the company so that it could not be kept would impair the value of the company and violate the fiduciary duty of the board.
The "business judgement rule" means that boards get a huge amount of latitude in how they run their business, especially absent "disloyalty" to the business (like enriching themselves at the company's expense).
In a 2005 case (re Walt Disney Co. Derivative Litigation), the Court said "not entitled to any remedy unless the transaction constitutes waste... [that is,] the exchange was so one-sided that no business person of ordinary, sound judgment could conclude that the corporation has received adequate consideration."
Your valuation has all sorts of uncertainties. Is there actually a market for the data? Even if it is from old, proprietary sensors? How comparable are those privacy experiments to a situation in which the sensor package is literally roaming your house 24/7? Is customer interest in privacy increasing (particularly in iRobot's target demographics?) and are there sales/reputational benefits for being one of the first companies to adopt this policy first? Any one of these is enough to ruin your case.
It would have to be something really egregious, like a contract in which the data transfers to the CEO's brother-in-law, for it to be a breach of duty.
* Business Judgement Rule: https://en.wikipedia.org/wiki/Business_judgment_rule
* Dodge v. Ford is where this idea of shareholder profits first comes from. However, look at the discussion under "Significance" https://en.wikipedia.org/wiki/Dodge_v._Ford_Motor_Co.#Signif...
* Article by a law prof at UCONN: https://medium.com/bull-market/there-is-no-effective-fiducia...
Directors are Directors because they represent a big chunk of stock. Billy Bob's VC invests in your company, gets a big chunk of stock and a seat on the board. Even 'outside' directors are often compensated for their time in stock. As a result every board member tends to make their judgments based on whether or not their stock value will go up or down.
Karl Ichan doesn't buy a bunch of stock and get himself on the board just for grins and giggles. No, he expects to wield the managerial power the board has over the CEO to enact policies that will improve the value of the stock he went out and purchased so that he can then sell it for a profit and move on to another company.
As a result discussions at the boardroom level are around "is this good for the stock or not" and in that vein I would do the analysis.
It seems quite interesting as a concept what a company is once it's bought and traded. What, exactly, is the precedence of interests? Is it state-first? Or owner-first? Or maybe a combination of the two?
... or am I just too drunk?
I am not an expert, I've played a part in no more than 10 companies being sold or purchased. In all the cases that I participated it the process involved creating a 'third' company that was essentially wholly owned by the acquiring company, and then that third company 'bought' some or all of the assets (physical or intellectual) of the company being bought. And then the purchased company was typically dissolved after a period of time that would allow for things like contracts or other actions to transpire while the entity existed. At a future point the intermediary company gets absorbed into its parent with a full transfer of all assets and that triggers its own dissolution clause.
The bylaws have always been the controlling documents for who could approve the sale, and how that process would complete. California adds a 'fairness' hearing sometimes[1] which I was told provides some protection against future lawsuits.
Bottom line a company is its employees, contracts, customers, physical property, intellectual property, cash accounts, and financial obligations. The entity or body authorized to sell the company can sell any or all of those pieces.
> Wouldn't it actually just deter companies interested in acquiring out of interest
> in the acquisition's customer data?
The long held theory here is that the share holder's elect board members to represent their interest in the company, which is notionally the 'value' of their shares, so for the board to be doing the shareholder's bidding they would not allow the CEO to explicitly damage the value of the company by impairing one of its assets with a destruct clause.
As others have pointed out, looking only at the costs associated with a decision to "spend" assets is not what fiduciary responsibility requires.
A purchaser looking to woo your clients (and continue to provide your service) rather than exploit your goodwill shouldn't have trouble retaining the majority of the data.
There's no reason to force data deletion, but leaving the choice to customers if they want to trust the purchasing company the way they trusted you prevents a lot of conflicts of interest and abusive behaviors.
I recently had a case with one such services. They store personal information that could be very damaging if it got out. SSN, driver's license, name, address, phone, credit card info, etc.
I decided I no-longer needed the service and so I asked to close the account and clear all of my data from their servers and backups. I was particularly interested in the information that could be used for nefarious purposes given a data breach or a sale of the business.
They refused and gave me some bullshit reason.
I moved immediately to arbitration per their own TOS. Within a couple of days they became very responsive and have, in theory, cleared all my info from their service.
And that's the issue. How do you know? They could have lied to me just to get me to drop arbitration. Hard to know how far to pursue something like this. No company should keep information on you that you do not authorize them to keep. A consumer ought to have the right to have any and all traces of their personal data erased from databases with absolute certainty. Privacy ought to supersede everything else.
What if you want Facebook to clear all data related to your account? I don't think that's possible in the US, is it?
And invariably, I hear about that exact data becoming exposed a few months later. It's extraordinarily frustrating that even when you articulate this viscous cycle to them, they still blindly push forward with the same efforts.
This is one of those cases where it is impossible to prove a negative -- one can never prove that the data you gave to Facebook doesn't still exist somewhere, even if they look for it everywhere they can think of. Similarly, the company you entered into arbitration with might have made a good-faith effort to delete your data, but they might not have fully succeeded.
Therefore, one must assume that any data given to any company might somehow survive even if the company attempts to delete it.
I mean hell, at my company we are putting in tools and methods that will allow us to very easily remove data with a single command from our systems should a customer/partner request their data deleted.
There is no way in hell they actually destroyed my sample or my data. They quite rightly knew that they didn't need to convince me of anything or try to hide the fact they were lying.
Having it enables getting hacked.
Not selling data enables not having it in the first place, which eliminates the entire failure class of "getting hacked".
NB, and not specific to you, but I'm increasingly cognizant of the general inability (or unwillingness) of people to follow chains of reasoning forward, or in this case and more readily, backwards, to identify likely consequences and/or causal / contributory mechanisms, in even only modestly complex cases.
It gives me a great deal of concern over the creation of an even only modestly complex information infrastructure.