Which seems kind of ironic, for Yahoo’s business is all about communications. But being able to channel communications isn’t at all the same thing as being able to communicate well.
By the way, for you movie fans, no, I’m not comparing either myself or Yahoo to a warden or a prisoner, but apart from the circumstances of Cool Hand Luke, this famous quote does seem to be very apt to the circumstances surrounding the Yahoo! Groups Neo roll-out.
Neo is a moving target (thank goodness, actually), so I’ve updated my prior post to strike some of the things that are no longer true. I was planning to write this post as a status update, “Twelve weeks of Neo,” but events got ahead of me and now it is fifteen weeks — a full three-and-a-half calendar months.
Why the silence?
One of the inspirations for this post was an October 30th sequence of conversations on Twitter involving Jeff Bonforte that inspired other thoughts, such as how the Groups team mishandled the communications about Neo as badly, if not worse, than they handled its technology. My prior post touched on the failures to communicate, but now I see there’s more to the story. Jeff is Senior Vice President of Communications Products at Yahoo!, a position that encompasses Groups, Mail, Messenger and other products, so it is probably safe to assume that he knows what is going on within Groups and why and how that fits in with Yahoo’s other products. If he doesn’t, well, then they have an internal failure to communicate to deal with.
One of the odd things about Yahoo’s communications on the subject of Neo is that there has been no statement in the name of the Yahoo! Groups Product Manager. There is such a person, Jeff told me so, but he hasn’t answered my direct request to name him or her. Other sources (GMF 48171) report that to be Hari Vasudev. I would have expected the PM to take a fairly hands-on role in such a major roll-out. Indeed, in the far more limited New Groups Calendar roll-out of 2012, the prior PM for Groups, Ashish Parnami, did engage one of the moderator tips and tricks groups, YGOG, to get early access to the new Calendar (message 41500). There are a couple things about that roll-out that are familiar though: It proceeded piecemeal and it didn’t stop, despite manifest defects and user community outrage due to its dysfunction and clumsy user interface.
I should add that even though I was “on point” for the user side of the calendar roll-out, as both moderator of YGOG and owner of YahooGroupedia, I never intended any of the criticism of the implementation or the handling of its roll-out as a personal affront to anyone at Yahoo, least of all Ashish, who I hoped would continue as an engaged PM. I would much prefer to have a good working relationship with the Groups team, top to bottom, so that issues can be discussed frankly. I try to make sure that my criticism is constructive, aimed at improving the Groups as a product and its user experience, so in that regard I find the silence of the current PM to be both odd and disappointing.
Why the lack of a pre-announcement or trial period?
A clue to this comes from Jeff’s comment when asked why there was no forewarning:
Given the history we were bunkered down.
I can’t know the full meaning of that, whether it is in reference to the Calendar roll-out or the earlier ill-fated “Facebook” and Beta roll-outs or to something else altogether, but it is a very good question. Could the Groups team have been so afraid of the user community’s reaction that they couldn’t bring themselves to set the stage? It seems hard to believe that, even harder to believe that they would think they could “slip this by” us. One possibility is that at some level Yahoo had a Potemkin village internally: Someone put up good demos of Neo and convinced higher management that the user community would accept Neo on its strengths. When Neo turned out to be incomplete and unstable in roll-out, Yahoo was then as stuck with it as we are and unable to concoct a rational explanation for how they got there.
Now I’ll grant that my explanation is fanciful: I’ve no evidence for it at all. But nature abhors a vacuum and human nature is no exception. Yahoo’s silence practically begs the users to fill the void with their darkest imaginings, and if you read the Feedback forum, they certainly have. Yahoo might not think so, but my explanation is relatively generous. Others posit that the CEO said “Everything must be modern, fresh and simplified. Now,” and that goal superseded all others — features, reliability, and the user community itself were sacrificed for that imperative.
What were the goals?
Here there is a succinct public statement:
Our aim in the redesign was to give the site a more modern, fresh and simplified layout, while maintaining the features that our users currently enjoy.
“Yahoo Groups Blog, 2013-08-29” (emphasis mine).
After fifteen weeks of roll-out and twelve-and-a-half weeks after that statement, they are still woefully short in reaching the goals they set for themselves.
“Modern, fresh and simplified” might be somewhat subjective. My own experience is that the interface works unreliably, and I’ve got up-to-date tools: IE 10, Firefox 25, and Chrome, running on a fully patched Win7 Pro, 64-bit. For others the experience has been far worse. Something that doesn’t operate smoothly, across platforms, simply doesn’t qualify in my mind as modern or fresh.
But far more serious is Yahoo’s failure to meet their secondary aim of maintaining the features of classic groups. Yes, Yahoo is making progress in bringing the functionality back, but that ought to have been a primary goal — it ought not to be still unmet fifteen weeks after roll-out began. The Feedback forum is replete with case after case of where it has taken a clamoring from the users to get a feature restored. That’s shameful — it shouldn’t take a popularity contest in a forum to restore base functionality. Base functionality should be a given.
It is as if the designers and implementers of the Neo interface had no or minimal experience with using or managing Yahoo Groups, or it seems that no one bothered to write down a comprehensive list of Yahoo Groups features and functions and then go make sure Neo implemented at least that much before roll-out. I’ll grant that compiling such a list would be no trivial task. I think very few people appreciate how sophisticated and subtle the operation of Yahoo Groups had been, with its interplay of features that met such a wide diversity of use cases, but the fact that it would have been a tedious effort does not change the fact that it should have been the necessary starting point.
So this is another issue in which some frank communication from Yahoo would go far in dispelling speculation and distrust. What were Yahoo’s expectations at the time of that announcement? They can’t possibly have thought they had achieved their aims, not given the volume of reports already present in the Feedback forum giving them list after list of missing features. Or were they so deafened by the cries of outrage that they couldn’t hear the rational voices?
So why roll it out while it was so manifestly unready?
Staring at tea leaves to divine the rational for a corporate decision isn’t necessarily a path to enlightenment, but given Yahoo’s failure to communicate on these issues, that’s what we’re left with. When asked directly, the answer was similar to the prior statement about why an opt-out couldn’t be allowed:
the old Groups runs on technology that is no longer remotely modern, which in tech causes tons of issues from QA to monitoring … it isn’t black and white, but we couldn’t stop groups. And time was ticking on the backend.
It is an answer that may be true, but doesn’t explain anything. Time is always ticking; that doesn’t really give a sense to the urgency. We are left to infer that they feared that a complete failure might occur before the roll-out could be completed unless they acted immediately, but that is an inference that doesn’t hold up under common-sense reasoning — not given that the Classic interface was still visibly in operation through Oct. 22nd when Yahoo pulled the plug on the moderation plug-in.
The what? One of the things Yahoo did right in this roll-out was to protect the operation of the many Freegle and Freecycle type groups which depended on Ed Hibbert’s Firefox plug-in to make these large, high-traffic groups practical to manage. When the Neo roll-out began, it was immediately realized (by the author and by the moderators of those groups) that the plugin, as it stood, could not be used with the Neo interface. So they petitioned Yahoo for remedy and they were granted a reprieve. Yahoo disclosed to the plugin’s author how to request that pages be delivered in classic format even when the user had been selected for the Neo roll-out. This let those groups continue their recycling efforts and gave the author time to raise funds and begin the implementation of a set of Neo-compatible moderation tools.
Not surprisingly, word of this “escape hatch” quickly spread via Yahoo Groups and in the Feedback forum itself. I give Yahoo credit for this: They stuck with it even though many criticized them for this tacit admission that the switch to Neo was not irreversible. Naturally, those that did not understand the real purpose for the reprieve were outraged when it was removed, but by that point Ed’s Neo-compatible tools were sufficiently developed that the affected communities could continue their operations. I don’t know how Yahoo’s management feels about this footnote in the Neo roll-out, but I salute those who made the decision to support those groups and those who stuck with it. I wish there were a way for Yahoo to claim that modicum of credit despite the howls of protest. It is another place where a bit of timely communication from Yahoo to the community at large might have headed off a lot of misunderstanding. Yes, calling attention to it at the time would have encouraged many to use the plugin who were not covered by its reason for existence. But that happened anyway, as it was surely bound to.
I don’t know how the saga of the plugin fits into the bigger picture of why Neo rolled out manifestly unready for prime time. Perhaps it doesn’t at all — an escape hatch for the plugin users would have been necessary even if Neo had every classic feature and had been reliable from day one of the roll-out. It might have been totally unnecessary if the roll-out had been announced in advance and users given a chance to kick the tires voluntarily, but that’s a different question.
Jeff’s reference to the ill-health of the classic Groups’ back-end database has spawned a lot of speculation about what that means exactly. There was one visible manifestation of trouble that had been lurking for years: the inability to keep the membership database for large groups (more than 400 users) in sync. And that problem had grown in the first half of 2013 to become one of the most frequent issues mentioned in the moderator groups and at Answers: Why does the member I just approved not appear in my membership list, or conversely, why does the member who unsubscribed still receive group messages? Almost by rote the Support Agents in Yahoo Answers were offering to re-sync the membership database to resolve these and other membership related issues. That’s a problem that should not be happening, and Jeff’s reference to QA and monitoring may relate directly to that and perhaps other issues of that type.
But to a layman’s understanding that problem doesn’t speak directly to an imperative to roll out a new system prematurely. This represents another missed opportunity for communication from Yahoo. Perhaps if they would offer a cogent explanation for the situation as it unfolded, they could have avoided an awful lot of ill-will.
What can Yahoo do about it now?
I’m sure that we’ve passed the point where Yahoo believes Neo to be solid enough that there’s no point in letting anyone opt out back to Classic. That point is hugely debatable, but it must be debated with facts. Mere clamoring for a return to the Classic format has fallen on deaf ears; only by making sure Yahoo knows just how much more work Neo needs is there any hope of convincing them that allowing an opt-out to Classic is a reasonable interim remediation. From my perspective that means documenting each missing feature and each bug or unreliable operation in the Feedback forum or with Customer Care directly.
One key difference in perspective here: I’m a moderator. My view of Neo versus Classic is heavily influenced by the usability and reliability of their respective moderator controls and features. So far, Classic continues to win hands-down by that measure. People who don’t moderate Yahoo Groups or whose groups use few of the features may not have the same view. I’m told that from the ordinary user’s point of view, Neo at fifteen weeks isn’t so bad. I’ll take that on advisement, but even if true, that’s no reason the rest of us should be denied remediation by way of an option to switch to Classic.
Yahoo’s response thus far is to say that they are focused on improvements to Neo. By inference, all we need to do is continue to point out its deficits and defects and then be patient while they work on it. In the first few weeks of the roll-out that answer might have held some sway, but after fifteen weeks of dysfunction, that claim has worn thin. “Soon,” as in “it will be better soon” has lost all meaning.
An option to use the Classic interface is something they could deliver nearly instantly. I say that based on the testimony from some U.S. users that they still see the classic interface for their U.S. based groups. If that is true, then whatever mechanism Yahoo set up to run Classic and Neo in parallel from Aug. 4th to Oct. 22nd is still functional.
All it would take, seemingly, is a menu item in the tools menu (gear icon, top right of every page) to select the other interface. It is true that some people would cling to that option and argue for its perpetuation indefinitely. Without an explanation as to why that is infeasible, they can only assume that Yahoo is being obstinate. I’m willing to accept the inevitability that the Classic interface must eventually be shut down for practical reasons. What I don’t accept, without better evidence, is that the day for that has come and gone, not while, in my own experience, Neo is still so badly inadequate. Fifteen weeks and counting. What if it is another fifteen weeks to get the serious issues finally put to rest?
If there is a technical reason that Yahoo can’t provide an option until then, they certainly haven’t communicated what that might be. If the reason is more one of goals, perception and management, then Yahoo should be forthright about that, too.