FriendFeed, the multi-network activity aggregator co-founded by GMail
creator Paul Buchheit, announced
today that it has entered the crowded field of real time search. FriendFeed was already the
best way to learn what early adopter social media users were saying about any topic across blogs,
Twitter, delicious and other diverse social media sites. If FriendFeed wants to step it up to the
next level and challenge business-class conversation trackers, we believe there are four steps
the company needs to take.
We think that would make a whole lot of sense. In fact we think that if real time search were
turned into a business tool it could challenge social media monitoring services like Radian6, Scout Labs and Sysomos. Here's what we think needs to happen in order for that to
become a possibility.
Sponsor
We already use FriendFeed to keep track of who quietly touches out blog posts out around the web.
For example, our recent post Google
Updates Blog Search - Where's The Innovation hasn't gotten any comments yet - but FriendFeed
shows us that leading marketing blogger Andy Beal
shared it with his network on Google Reader. That's good to know.
We think FriendFeed could offer some of the most sophisticated social media conversation tracking
on the web, if it just took a few steps in particular.

Broaden the Index Beyond Opt-In
Right now FriendFeed tracks what users say and do across more than 40 different social media
sites and any RSS feeds (like blog feeds) that users input as part of their profiles. It's a
great way to track people and topics on networks you yourself don't participate on.
Because FriendFeed is such a high-profile startup, many people have set up accounts just to try
it and have their activities pulled into the site automatically even though they no longer use
FriendFeed itself. That adds to the richness of the site's search function.
If FriendFeed wants to offer full-service conversation tracking, though, it is going to need to
go beyond the early-adopter crowd that has opted in to having their activities imported into the
site. FriendFeed is going to need to proactively discover and import feeds from users of Twitter,
Delicious, SlideShare, BrightKite,etc. and bloggers who have not set up FriendFeed accounts. This
will increase the usefulness of the site's search function by orders of magnitude.
That's no small task! Many startups have tried to do social-media-wide search in the past but few
can achieve the scale and speed needed to pull it off well. Two things make us think FriendFeed
can do it. First, who better than the creator of Gmail to achieve new hights in rapid, scalable
information delivery? Buchheit isn't the only former Googler on the team, either. Second,
FriendFeed has been engineered from the start to import massive amounts of data. A number of the
streams FriendFeed pulls in aren't even from RSS feeds as you'd expect, the company's co-founders
told
ReadWriteWeb in an early interview that they have built to import from a wide variety of
Application Programming Interfaces (APIs) beyond RSS.
Spam Control
Right now there's a fair amount of spam on FriendFeed and we imagine it's only going to get
worse. There's also a lot of people who use the service as their RSS reader - so blog posts and
search results show up in your search results even though no one has touched them.
This wouldn't be a difficult problem to solve. FriendFeed is just a few steps away already from
allowing searchers to query all sources except for manually imported RSS feeds except when a feed
item has one or more comment or "like" added by a user.
Apply Business Savvy
There are all kinds of ways that FriendFeed could become more business savvy. One way we would
suggest is through making more use of LinkedIn.
FriendFeed users were able to associate their LinkedIn profiles with their FriendFeed accounts
since the start of the service and job changes used to be displayed right along with Tweets and
other online activity. It was great. Unfortunately, LinkedIn knows what a pot of gold it sits on
and took the noxious step of cutting off these kinds of importing functions by obscuring the HTML
on its profile pages. FriendFeed was scraping those pages for changes and it was a great service
for everyone. It was pure folly by LinkedIn, it wasn't specifically targeting FriendFeed, but as
a result FriendFeed users no longer see when their friends change jobs and click through to
LinkedIn to learn more.
Fortunately FriendFeed hasn't removed LinkedIn as a field that can be viewed for users, it just
doesn't update anymore. When you see that someone has said something in your search stream you
can many times click through to their LinkedIn profile to see what they do for a living and what
their job title is. FriendFeed could display job titles by default on a business version of
FriendFeed if LinkedIn was more agreeable, or FriendFeed could look to the much friendlier and
social media savvy Google Profiles instead.
Knowing the job titles of people who have bookmarked your web page in Delicious or shared it in
Google Reader would be really valuable.
Some other business oriented rules, like alerts when certain discussion thresholds are past,
would go a long way too.
Bring Back Aggregate Analysis
When FriendFeed launched it offered some great data visualization showing which other users you
"liked" content from the most and who "liked" your content the most. It showed which services
most of your content came in through in a pie chart.
Unfortunately in a recent redesign aimed to make the service more mainstream-user friendly those
visualizations were eliminated.
Bring that and more back and you've got a viable competitor for services that businesses pay
hundreds or thousands of dollars a year for. Add some sentiment analysis, made easier by
FriendFeed's "like" feedback function, and you've really got a desirable product.
Will FriendFeed take these steps though? That depends on whether it continues
its Quixotic quest to capture more everyday consumer users for a cross-network, real-time
conversation aggregator (!) or whether it finds audiences that appreciate its value and
start building out features that they will pay for.
Discuss
