Archive for the ‘site optimization’ Category
Mahalo: Yikes…that’s product marketing and data analysis?
Update: CEO Jason Calacanis was nice enough to respond to our post below:
His comments:
fyi: we have a user lab at mahalo at did 75 hours of testing over the last year. All of our site changes are considered deeply with a combination of user testing, user feedback, and metrics testing.
In terms of who’s responding there are normal folk in there, user experience folks, and designers. I would go read the 200+ comments and 50+ notes on the image a day later. Some are very considered, and the feedback while not a perfect roadmap is very helpful.
Talking to users is NEVER a bad idea. Listening to everything they say? well, obviously you don’t want to do taht or you have chaos.
best jason
Jason, thanks kindly for responding to this blog; it is well appreciated. I agree talking to users is never a bad idea; I said as much in my customer service as the new marketing post.
And I agree that if you get user feedback from UI folks and designers, it is priceless. So I will reflect that in the post.
However, two things:
1) I think there is a user fatigue that comes with too much messaging. If everyone twitter’d or flickr’d all their site designs a user would become deluged with noise.
2) You have to ask or beg the question of the feedback: How often do you use Maholo? If not the feedback itself is noise.
My conclusions:
- Talking to users is never a bad idea
- Processing their feedback correctly for the product roadmap is crucial
- Who is the target audience for Mahalo and have you reached out to them effectively through social media circles?
Thanks so much for responding.
——
I’ve commented before on Mahalo.
Upon a post by their founder on his blog, I am exceedingly worried for their investors.
The title of the post is Social Media Focus Groups as a value for Twitter and other quick-response, signal-frequency services. Most who read this blog know I feel the value of Twitter is in labeling the data with statistics for personal review and reducing the noise.
In terms of the blog post referred here, I couldn’t disagree more on both the use for Twitter and the explanation for the social media focus group.
First, on the social media focus groups, anyone who does online site optimization knows that what users do is very different then what users say. That is nothing new.
Second, on using Twitter and other such services to gain feedback, probably best to use this avenue to comment on large news than rudimentary page designs lest people stop paying attention to your posts. Do I really care if I use a service whether some guide thing is on the left or some links are on the right? Probably not, so I would label that message as spam going forward. If you released a new service that added value, told me about it that way, and I found it immediate value, sure I would be accepting of that message.
Further, as I’ve talked about in past posts, “who” is responding is just as important as what they do. The folks responding to a little usability focus groups on Twitter or even Flickr are probably not the audience that you need to gain and or optimize for to increase growth. Except if these users are UI or design experts that are offering feedback (thank you for the correction).
Banner Ad is to Advertising Campaign as Twitter is to….
….personal news portal! Personal news portal!
TechCrunch, again trumping others, is carrying a story that Twitter is testing instream advertising.
All I can say to this is, “No! No! No! Look at the opportunity.”
My last and only post on Twitter merely talked about the power in building signals.
(I’ll have another post coming shortly on the signal-receiver relationship online).
Twitter is missing a golden opportunity here to build a destination site.
Much like banner ad campaigns are distributed widely online and lead users back to a Web site is the relationship that Twitter should have between their users and a destination site. In essence a user has opted in for the most powerful advertising perhaps ever: a relevant content message from a trusted voice that gets pushed to a device that is always on them.
This permission-based content push is the core competency and value of Twitter and they should not compromise by introducing noise. This would corrupt the system completely. Suddenly, that desire to view a message would be caveated with, “Wait, maybe it’s an advertisement.”
Twitter should provide a web-based service that then capitalizes on their core competency.
Give me statistics on messages and wrap marketing around this.
Introduce relevant content to twitter categories or my grouping (either automated or with manual editorialship) and wrap advertising around this.
Provide a customized newsletter feed from my twittering and sell sponsorships.
Review point: Google introduced search ad campaigns around their core competency (their natural results) not in them.
—-
More comments inline with this sentiment on Mashable today.
Cluuz: Like it, but not really next gen guys…
I’m more poking fun at Cluuz. I like their “array” display of search engine results that allows the user to select an appropriate result by “promiximity” to what their looking for.
http://altsearchengines.com/2008/04/07/the-next-generation-search-engine-cluuz/
For those that have been around awhile, we rememember Gnod’s Music and their Music Map. Still a very effective recommendation for music which originated in the 1990s.
Are you a Radiohead fan? http://www.music-map.com/radiohead.html
Search engine interfaces are a lot like multi variate testing, you never know what comvination of features and visuals will illicit positive feedback and return usage.
I’m also looking forward to the debut of PowerSet this week.
Newspapers: Step IV: Identify silos of initial value
We have the CTO.
We’ve talked to our advertisers (and set some expectations)
We’ve improved our reporting.
Step IV:
Identify silos of initial value
Summary: I’m going to make an assumption here that a newspaper site does not have enough content producers, product managers and developers (even with the new CTO) to solve everything at once. I think this is a fair assumption.
Therefore, we’ve have to focus on one vertical or set of advertisers and solve their needs. The goal is the following: increase all metrics of a set group of advertisers or vertical across all points of the supply chain.
Directive: Solve a vertical.
Let’s use the travel vertical example. We have 7 travel partners that really like our attention to their campaign. Next they tell us what performs best from them (along with the accompanying metrics — brand and direct response). We investigate where they are displayed and we investigate how users are getting to these pages.
We form a task force. One that has carte blanche to: a) not seek approval on all content issues and b) ability to prioritize work for anyone as they see fit.
(Ideally we have a strong product manager that has followed the CTO leading the charge.)
We do an extensive competitive analysis, we identify success metrics, we build or repackage content as necessary and we strike distribution deals as quickly as possible regardless of price. (Price will just slow us down; we can always renegotiate later).
All the while, we inventory the effort level and manhours needed to achieve this. We identify where processes are repeatable for other veriticals or advertiser niches.
What do we have? Well after a few iterative takes, we probably arrive at something of value based upon a valuable advertiser metric (some examples in travel could be: impressions, clicks, bookings, viewthus, email sign-ups or other).
So know we have a success. We also have the manpower assessment it took to complete this (our operating cost). Further, we have process points to improve operations.
Thus we can do this again and again for a series of clients. This isn’t the end game mind you, but it solves some of the following:
- advertiser value
- advertiser expectation and customer service happiness
- content / production costs
- revenue
Once, we have more users, we can make more valid assumptions and try some riskier, highever value endeavors.
While this post may seem obvious to any user, I wonder how many newspaper sites are evaluating their effort level based upon a specific ROI and putting in processes to improve the profitability.
Newspapers: Step III: It’s all about events and reporting
Okay, I now have a CTO, or I’m in the process of getting one, and I have feedback from my advertisers. I have momentum.
I have goodwill with my advertisers becuase I listened to their concerns and I showed them how serious I was by getting a CTO and vowing to ramp up my technical capabilities.
What next?
Step III
Evaluate what reporting I have and what I need
Summary: Many online publishing companies incorporate a webtrends or an ominture and there done with it. Essentially these publishers are letting reporting software companies determine their metrics. It’s a travesty, it happens all the time, it’s incorrect and there it really no excuse for doing it.
I’m not saying get rid of your Ominture or Webtrends service by any means. I am suggesting that if you have the desire to be a big publication that you need to manage your reporting interface so that you can define new “things” that mean value.
For example, maybe the “weather” page is the largest entry page on my publication site, well maybe I make it a defined goal to improve the content and navigation around weather.
Regardless of “what” it is you need to measure it.
And rigid reporting structures don’t allow you to do this.
You need a flexible event model that allows you to identify metrics as they happen.
Do you have this?
As an additional note, a reporting web display that makes sense for your business drives sound business decisions. I can’t tell you how important that is.
Directive: Easy one here.
- Understand what you have (internal, external, what’s the support level). Have you mastered Omniture, can you dynamically change reports and interfaces.
- Evaluate what you need. You have this already from you advertisers, we don’t have it from content (but that’s okay they don’t know what they are doing anyway). And now your CTO will help you understand what’s supported and what’s not.
Reporting is a “product” in an of itself for any major online publisher or advertiser. It is living and breathing and it needs constant product development and support. It is part and parcel to your content.
Create the roadmap for it and start executing now.
OMMA: What’s next, a story on the X10 camera? Ridiculous
In Twitter-like fashion here:
- Get mail yesterday
- Flip through to magazines
- See OMMA with a story on performance marketing. Omma, okay, cool.
- Flip to the lead story: “Click me baby one more time”
http://publications.mediapost.com/?fuseaction=Articles.showArticleHomePage&art_aid=79703
- Third sentence: “For almost a decade, performance marketers have placed so much importance on initial click-through rates that they’ve forgotten nearly everything else.”
- Close magazine, start crying, curl-up in a fetal position
- Get up go to my computer, log onto Prodigy and dial up Preview Travel for some airline tickets and then it’s over to Infoseek for my financial news…wait I can go to RagingBull, great!
Did anyone else feel that OMMA made a mistake and reprinted a turn-back-the-clock issue from 2002?
Hey Netflix, are you just looking at the click. Is that how you scaled your business? What about you MyCokeRewards? Omniture has about $143M reasons to tell you why it’s not just about the click. Oh yeah, and they’ve been in business for awhile.
Anyone using Google Analytics? Or even a BuzzLogic?
It’s a sad day when an online publication so widely followed and accepted by the luminaries in the media space puts out such a patently late piece and calls it news.
Maybe that’s why there are blogs…..thankfully.
Leave a Comment
Comments (1)
Comments (2)