"The press are now all over the OFA team and criticising their ability to procure technical solutions or web services (rightly or wrongly). But in this case, obviously we’re a third party service provider so this really has nothing to do with the President’s team’s ability to execute."
It is indeed a reflection of the President's team's ability to execute when they reach out to a vendor on a Friday for a Monday call, mangle the details so that the vendor and the audience have mis-set expectations, and, most importantly, did not do testing to determine whether or not their event will reflect well on their boss and organization.
They've done these calls before - for example, after the State of the Union: http://www.politico.com/politico44/2013/02/obama-doing-ofa-c... -- so dropping this order in the lap of a new vendor with 72 hours to go almost guarantees that some mistake will happen.
I understand this is OFA, not the White House, and perhaps for a seat-of-the-pants campaign style event under normal circumstances, this would be excusable.
But these are not normal circumstances.
When your boss is under fire for his rollout of the most important government tech program since the Moon landing, and you arrange a call to discuss that rollout, it is imperative that you make sure the technology works. Guaranteed. No questions asked.
If you can't guarantee that, then don't use that vendor, don't do the event as a conference call, or don't do the event.
Why? The resulting PR mess, which might not be germane to a start-up tech audience, but is very germane to the Office of The President, will come and will be costly.
The post serves to reinforce the image of a management team around the President that does not understand that making very urgent requests of technologists does not mean that the technology will automatically obey.
They do maintain some sort of vain presence, so as to appease the college crowd that make use of large amounts of social media, but no actual debate or interesting conversation has been inspired by it.
The AMA on Reddit in particular was one of the most pathetic PR stunts I've witnessed in a while. A few generic responses to mundane and chauvinistic questions. But, since his campaign team ended up referencing the "NOT BAD" meme, Reddit swallowed it all up with gusto.
It was sad to watch.
If you honestly expected any candid answers from a sitting President that diverged from preset soundbites, I have a bridge to sell you.
>Unfortunately the call was marketed and promoted as a “conference call” ... which does happen to be what Mixlr is intended for
Okay, so what is Mixlr for? I poked around and found out briefly, but I'm still curious what Mixlr allows that they didn't use and could have. What did you mean by "nor to the extent of what Mixlr allows"?
>This led to some pretty mis-set expectations and disappointed users which wasn’t a great tone to recover from.
How do you know users were disappointed? Do you have any data, or at least anecdotes, that you could show? Do users rate their experience, for example? How did you recover? And how did you decide that you recovered already?
>Even 10 trolls are a lot louder than 30,000 supporters
What actually happened? What features enabled a few people to effect the experience of the other listeners?
>Note to self... disable comments
I gather that disabling comments is not a feature available to the broadcaster? Maybe it should be; are you considering adding it? What other changes will you be making as a result of this experience? How will you "do better next time"?
>the press are now all over the OFA team
The write up makes it sound like the whole event went smoothly, yet you're also saying that there are negative reports in the press. I read the DailyMail article, and it sounds possible that they are exaggerating by citing the minority of users who couldn't connect. What is your takeaway here? Do you have to be perfect in such high profile situations? Or are a few errors expected? Are you getting flak or losing credibility because of the press, or is it all falling on the OFA team?
> the call was flawless for the vast majority of those listeners for more than 99% of the call
Are more details available? Maybe it's just me, but this sounds wishy-washy (99% always sounds to me like "a lot, we swear!")
> we think we can do even better next time
Great! Is there going to be a next time? And, more generally, have you noticed any effect on the call on your user base?
I don't see what the incentive for NYT to overstate or mischaracterize things, unless just to cast Obama further into a negative light, so I'm tempted to believe what the press says over what a self-motivated startup CEO would write. The NYT article makes no references to trolls, but rather what seems to be valid claims from would-be listeners that they can't access the audio stream.
If the audio problems were truly experienced by the insignificant minority, and a small number of trolls caused the majority of problems, then why was the article written in the NYT as if the glitches were the primary story, rather than the content of the broadcast (which is not even covered in the NYT article). The truth is somewhere between the NYT article and Mixlr blog post, but most likely weighted more to NYT's side.
I think Mixlr should thank their lucky stars that the blame was placed on OFA and not on them. In fact, I would be thankful the Mixlr name didn't even appear in the article. Had they not written the blog post, no one would have associated the event with their service. Having the president use their service serves as a good reference, but this one teleconference would not be what I would consider to be a positive testimonial.
Lets assume that there were 100x that many failures, that would be 700 who failed, and 139,300 who could hear, which matches Mixlr's description fairly accurately. They had a large event that went well overall, but the problems that did occur caused a bit of a stir.
Frankly, I have never heard of these guys before, but they earned my respect via their post because they did accept their own accountability for the issues. They are not passing the buck, they acknowledged that there were problems, and they intend to do better next time. That is absolutely the correct attitude.
I agree with the parent poster that the truth probably lies somewhere in between, but like you, I don't think we can give NYT much extra credit here.
I'm also surprised to see that more technologists aren't empathetic to the "tens of thousands of end users" matter. In a group of people over 100,000, we can begin to expect a fairly sizable number of users who have configuration issues that would cause their own host of problems. I'm not indemnifying Mixlr, but it's worth keeping in mind when finding the middle ground about what really happened.
I think you mean "...because advertising".
Perhaps the NYT blogger's editorial goal is to counteract the blame aimed at OFA / Obama by pointing it at Mixlr instead.
Yup. Huge opportunity missed - they should have hired a real writer to tell the story, instead of "this happened, then this happened, then ..."
A "real writer" would have said the same thing but taken 5000 words to say it. It would be more fun to read, sure, but would take much longer to get the gist, and half of it would probably be made up to sound good.
"Unfortunately the call was marketed and promoted as a “conference call” and questions were invited in advance [...] which does happen to be what Mixlr is intended for [...] However, in the end they didn’t really use it that way nor to the extent of what Mixlr allows and it was just a one-way broadcast." (from the eighth paragraph in)
I'd also raise the issue of whether the $9.99 all-you-can-eat plan should have a specific limit in terms of absolute number of live connections. From their pricing page, it looks like OFA could have just used the Basic (free) plan. Even still, they will never recoup their costs from this one call even if OFA remains a $9.99/month subscriber for a decade. If "We worked all weekend..." means two engineers worked 4 hours Saturday and 4 hours Sunday, that's 8 hours/engineer * 2 engineers * $100/hour = $1600 of time was spent. That's 13 years at $10/month, not even factoring in infrastructure or bandwidth costs from the call itself. And I'm guessing that estimate is an order of magnitude less than the actual time spent. If I were a potential customer evaluating providers, I would be scared off by your current offer of what appears to be unlimited usage and unlimited priority support for $10/month as I'd have serious doubts of your ability to remain in business a year from now.
No doubt the effort invested in preparation for the OFA call will pay dividends for future scalability, but I doubt the people at OFA would have blinked an eye if you told them that there's a $5 CPM rate for connections beyond the first 100. 140K connections would have been a paltry $700. Obviously, only an analysis of your customer usage patterns would indicate whether you'd impact any serious customers if the basic plan was capped at 100 simultaneous live listeners, with additional tiers for higher peak use. Check out PubNub's pricing model for peak connections as this model makes financial sense: http://www.pubnub.com/pricing
This was such an outlier in terms of standard use cases of their system so far, I think the way they responded was great. There's little value in going back and saying "Ohh... this will cost $700, but it's not written on the website, is that still ok?" and risk dealing with rejection.
The value of having POTUS use your service far outweighs the engineering time spent on this. This is now a fantastic leading story for all sorts of marketing efforts and meetings moving forwards, such as this blog post.
Calculating the return on this effort through short-term ROI measured in days seems a bit shortsighted.
Congrats guys, it has to feel great having something like this go so well!
Maybe there's a number in the middle, but I think it's hard to draw a line. At the end of the day, if I were a US citizen, I would prefer my government to attempt with a $9.99 service that went almost flawlessly, rather than ditch it just because "it's cheap, it can't work, we must pay at least $X".
1. Mixlr in the header points me to blog.mixlr.com.
2. Footer links, all of them link to the page I'm on...
I know there is a link in the content of the post, but what about other posts? Why is there not an easy way for me to get to your main site without manually modifying the URL in the address bar?
How many times do you think the president needs to say "Sorry about the technical difficulties, the company we were using was having some troubles" before it hurts them?
Do you think more people are going to read their blog post here, or if the president was to tweet something?
The only thing not included is how long they are a member or what they pay, though the latter can be assumed since there are only two tiers of members (free and $10 a month).
> On Friday, one of their representatives dropped us an email
If you work at Mixlr I recommend you kill this comment. It's not best practice to compare your "pedestrian" customer to someone "really popular".