Point Solutions vs End to End Solutions

« Should WiFi Be Public Infrastructure? (continued) | Main | VC Cliche of the Week »

Point Solutions vs End to End Solutions

I wrote this on the flight out to the Web 2.0 conference. I hope it stimulates some interesting conversations.

---------------------------------------

One of the central tenets of Web 2.0 thinking is that lightweight "point solutions" that can be stitched together by the consumer are preferable to end to end solutions that are stitched together by the service provider.

I had a conversation with veteran web entrepreneur who challenged that assumption yesterday.

His view, one that I have to admit made me sit back and think, is that the early adopters (geeks) of the web 2.0 world prefer to stitch together point solutions, but that the mainstream web user will prefer an end to end solution.

Further, his view is that the leading portals; Google (I know they aren't a portal, they are a starbucks store), Yahoo!, MSN, AOL, IAC, etc will stitch together end to end solutions with a combined build and buy strategy.

And it follows from this analysis, that the portals will end up winning the masses and leaving the point solutions either as assets to be purchased or to waste away and die.

Obviously we need to figure out the answer to this question because we either follow the lead of Flickr and look to the portals as our exit strategy or we invest in and build these "point solutions" into viable long term businesses.

I don't think we have to make this call just yet and we can invest for growth without limiting the option of a sale scenario.

But this is a long term strategic question that everyone starting and investing in web services businesses needs to get a handle on if they want to make the exercise pay off at the end of the day.

Here are a some data points to think about.

Flickr:
Flickr hasn't improved in any measurable way now that it is part of Yahoo!. In fact, I am finding it to be buggier and flakier than usual lately. Probably due to scaling issues, but being part of Yahoo! doesn't seem to be a magic bullet in that regard.

And some of Flickr's users are smarting from being forced into Yahoo!'s registration scheme.

There hasn't been any noticeable integration of search, groups, 360, or MyWeb with Flickr yet. Yet is probably the operative word and we have to watch this closely to see if these integrations, which must be coming, improve the utility and vitality of the Flickr experience.

As a fanatic Flickr user, I tend to doubt it. I love Flickr for what it does for me today. I have done the integration that I want, with my Treo, with my blog, with delicious. I doubt that Yahoo is going to give me anything more than what I already get from Flickr.

AOL vs. Google:
AOL is the classic end to end solution. They built everything inside their walled garden and you can't integrate any web service with AOL. Of course that is changing, about 10 years after it should have. But you know what I mean.

Google is the classic point solution. They built the very best web search engine.

Which would you rather own?

Blogging Platforms:
Each of the major portals has a blogging platform, Blogger, 360, MSN Spaces, etc.

I don't see any of them taking a cut out of Live Journal, Xanga, Typepad, Wordpress, or any of the "point solutions" in the blogging world. In fact, the platform with the most mojo these days seems to be MySpace.

News Corp doesn't have any web services to integrate with MySpace and I suspect they are going to get a fantastic return on this acquisition without any "synergy".

Craig's List:
I just shake my head every time I think about Craig's List. It has existed as an island all to itself for years and just gets stronger and stronger.

The best web 2.0 mashup I have seen is Google Maps/Craig's List.

Google is going to launch a Craig's List killer sometime soon. Let's see how it does. I have my doubts that it will do much more than Yahoo!'s assault on eBay in the Web 1.0 days.

Wikipedia:
This is my favorite of them all. The same web entrepreneur who got me thinking about this issue suggested that a better wikipedia might be built with a more commercial model. No way. Wikipedia is the bomb.

I read a post today on the Socialtext weblog about the way Wikipedia is slowly taking over Google's page rank algorithm.

Think about that. No SEO or SEM going on there. Just lots of link love and internal linking and voila, wikipedia slowly takes over Google's page rank.

That's a jujitsu move if I've ever seen one.

The fact is that these "point solutions" have a vitality that comes from their authenticity, their simplicity, and their sizable and active user bases.

I heard a user of one of these services talk a couple weeks ago about the "emotional connection" he had to one of these services.

I know what he is talking about. I feel that way about Flickr, Typepad, Google, My Yahoo!, delicious, and many other web services I use on a daily basis.

I have rolled my own web experience and it is unique to me. It is mine.

Yahoo!, Google, MAN, AOL, Ask, and anyone else who wants to try is not going to take it away from me.

And its people like me who make this whole Web 2.0 world so vital. We are the content creators. Without us, the services will be empty.

Try my Web 2.0 for example.

The launch of Marc Andreesen 's Ning yesterday makes me think that its quickly going to get easier to build new web services and that we'll see an explosion of these point solutions over the next ten years which will be even smaller and more lightweight.

So my thesis is that end to end solutions aren't going to work as well as highly integrated but separate services that build and hold emotional connections with their key users.

And those users will provide the lasting value.

But that is just a thesis. I want to test it. By blogging it and soliciting comments, links, trackbacks, etc. And by asking everyone I know what they think. And by bringing really smart people together to debate issues like this.

As I learn more and develop this thesis or debunk it, I will keep you posted as always.

October 5, 2005 Venture Capital and Technology | Comments (12) | TrackBack (10)

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/t/trackback/5934/3310004

Listed below are links to weblogs that reference Point Solutions vs End to End Solutions:

» Developing with Ning from The Community Engine Blog
Fred Wilson thinks Ning might lower the cost of entry for creating web applications designed to fulfill a specific need very well. Based on my experience, I don't think so, but I have a suggestion which might help. [Read More]

Tracked on Oct 5, 2005 1:47:59 PM

» Do the Giants Win? from Blogspotting
Fred Wilson asks whether point services (think Flickr) or end to end ones (think portals) are going to win the masses. The question strikes a chord with me because just last week I was wondering about the future of one of the company's he's invested in. [Read More]

Tracked on Oct 5, 2005 5:25:31 PM

» Chunky or smooth from SolutionJunkie -- Doug Giuliana
While end to end solutions may not work well for Fred, I do think there is a huge population out there that wants and needs some aggregation. Point solutions also suffer from being very specific, by their very nature. Look at what AOL is doing with s... [Read More]

Tracked on Oct 5, 2005 6:10:17 PM

» Web 2.0: heads or (long) tails from Mark Sigal's Blog - The Network Garden
Fred Wilson of A VC has written a really well articulated post called, Point Solutions vs. End to End Solutions that raises the question of whether the Web 2.0 mantra of lightweight point solutions that can be stitched together by [Read More]

Tracked on Oct 6, 2005 5:50:33 AM

» The classic question: end-to-end or point solution? from Raganwald
...That doesn't mean making a web service that can't do anything until someone mashes it up with another tool. You'll notice that point solutions are still solutions. You still have to solve one problem and solve it well. [Read More]

Tracked on Oct 6, 2005 6:38:04 AM

» ON MAINSTREAM USERS COPING WITH WEB 2.0 from *michael parekh on IT*
YOU AIN'T SEEN NOTHIN' YET Seeing the flurry of companies making their presentations at the overflowing and well-covered Web 2.0 Conference (my earlier post on this here), it's becoming clearer to many more that the the traditional ways that mainstream... [Read More]

Tracked on Oct 6, 2005 5:10:32 PM

» Web 2.0 This Week (October 2 - 8) from TechCrunch
Web 2.0 This Week October 2 - 8 What a week! Web 2.0 was absolutely terrific. There were hundreds of smart and interesting people milling about and cross pollinating their ideas. Our focus was on the new companies, of course, and we briefly wrote abo... [Read More]

Tracked on Oct 10, 2005 3:33:39 AM

Tracked on Oct 10, 2005 3:35:39 AM

» 新年の挨拶に代えて：マイクロソフトの戦略転換とRay Ozzieのメモ from 渡辺聡・情報化社会の航海図
本連載もあと四半期で2周年となります。何気なく始まったところから、気がつけば思わぬ長丁場になってしまいました。これからもこれまでと変わらず、世にただ流されるのではなく、一本筋の通った視点を届けるべく綴っていきます。ご愛顧頂いているみなさま、日頃お世話になっているみなさま、本年も宜しくお願いいたします。 ◇　◇　◇ さて、どこから始めたものかと仕事頭への切り替え準備で幾つかのサイトを見ていると、昨年話題を読んだマイクロソフトのRay Ozzie（Groov... [Read More]

Tracked on Jan 4, 2006 9:18:08 AM

Tracked on Jan 4, 2006 9:19:05 AM

Comments

You may find some reasons of the non-progress of Flickr since Yahoo acquisition in the post of this morning on flickr's blog

"Well now they are our corporate kin and we're squealing with joy (seriously - the people in the cubes NEXT TO US *HATE* US). "

The way it's written doesn't really make it sound like a joke.

check out http://blog.flickr.com/flickrblog/2005/10/upcoming_welcom.html

Posted by: d.durand | Oct 5, 2005 10:19:13 AM

Shouldn't that be "Wikipedia is 'da' bomb"? Seriously, though, great post. There's a reason that there are so many Flickr addicts and no (none that I'm aware of) 360 addicts. The points are pure and authentic and draw people to them in a way that a wholly-structured environment that shunts you here and there can't. That said, it seems impossible to keep any of these points (the successful ones anyway) free of the portals. Enjoy Web2.0, wish I could be there.

Posted by: Greg | Oct 5, 2005 10:45:01 AM

I love the "roll your own web" idea, Fred. It's something I've done, too (as has every other Web 2.0 adopter to date), but I do see a problem when I think about my mother doing it.

Maybe if her Mac helped her a bit more, but we all know how the big HW/SW vendors (i.e. MS and Apple) couple their products with their versions of services (web 2.0 and otherwise).

My mom still thinks that when i send her a link to a picture that I've put it on her computer. So we've got to figure out that layer of abstraction and somehow make a pluggable framework that is as simple as anything else that is very well done (i.e. iTunes, google, & c).

So that does seem to point to a "portal" being the actor who can pull it off. But, and it's a big but, this will lead to lots of peopel using services that are just "good enough", and not integrating the "best of breed" (ugh, I hate that word) services into their own Web 2.0 experience.

I'm fanatic about Flickr, despise 360, thing Yahoo Mail is just so-so, Yahoo search is just OK -- so I'd be stuck in a mediocre world were I to stick to only Yahoo's services.

Some company's going to have to make an OS, "webOS" or not (MacOS, Windows, a Linux distro), that works like this: one is simply offered some big juicy icons that represent each of these services, and when you drag them into your desktop, they seamlessly "install" and automatically know about and speak to one another -- regardless of which you pick.

Posted by: scott partee | Oct 5, 2005 12:47:50 PM

end-to-end solutions are better than point solutions when they are well-designed and the user experience is holistic.

however, if they are poorly designed, then by definition the point solution will be simpler, better understood, and likely will have greater adoption.

for example: our goal with SimplyHired is not to end in just a point solution for a great set of job search listings (altho that IS a core function & minimal rqmt), but to also provide related mini-apps and services that a job seeker would actually use during the job search process (location & commute mapping for jobs, salary research, company lookup, social networking contacts via LinkedIn, note-taking capability across various job boards & sites, etc).

if we fail to do this well or in an easy to understand way, then simply doing vertical job search will probably be better for the user.

however, if we pull it off the experience should be much stickier than a point solution which requires jumping to other sites/services to complete the overall process.

this is actually the promise of Web 2.0 sites & services: build your core featureset well, then use lightweight Web 2.0-style mashup integration with other related services that map well (no pun intended) to your target audience and their specific set of problems.

to summarize: end-to-end solutions are better if (and ONLY if) they 1) work well and 2) solve an overall process or problem, for 3) a targeted set of users with shared characteristics & issues.

point solutions work better if developers fail to build a good UI, try and bite off too big a chunk of the problem set, or try and group users with disparate needs & behaviors.

Posted by: Dave McClure | Oct 5, 2005 2:54:31 PM

Fred I disagree with you on Flickr.

http://thomashawk.com/2005/10/fred-wilson-is-wrong-about-flickr.html

Posted by: Thomas Hawk | Oct 5, 2005 4:11:26 PM

I like your thinking. What about this?

Actually, Wikipedia is a social site. Some
of the people who post there (see?) do
physics or chemistry and are very factual,
scientific people. And their friends DO go
look at their entries, either the ones with
their names on them, or the ones they
write.

And ... I know of a few writers (profes-
sional writers) that have arranged to
have their stories told (so to say)
exactly the way they wanted them to be
told. So now they visit and check on
changes, but of course some of their
friends and otherwise have changed things,
so they visit again. Finally, the
community stabilizes. But, no, it
doesn't, requiring more visits and more.

I could go on. I hope you see. Culture is a
compromise. Wikipedia just accidentally
became, well, a cultural center much more
interesting than the worker bees expected.

Because Wikipedia is a carnival as much
as it is an encyclopedia. Involvement.
You know.

sabadash

Posted by: "-" | Oct 5, 2005 9:27:59 PM

you're right ... A useful analogy is to break the systems we're trying to mash together down futher ... you'll find an SOA ... or Service Oriented Architecture. P2P is better that E2E for the exact same reason that SOA is better than Client / Server.

By now, SOA's are proven more efficient, reliable, extensible, robust. Why isn't this wisdom applied at 30,000ft? With API's and other feeds, there's no need to own flickr to realize its synergies. What will happen is that unrelated historical bias' within the acquiring company will garuantee that the acquisition under-recovers relative to its potential.

Things go wrong when when you stop listening to the geeks ...

Posted by: David Gibbons | Oct 5, 2005 10:28:38 PM

I have been racking my brain trying to figure out how web 2.0 will work in the Enterprise software market. We have been doing some cool Ajax stuff, linking to things like flikr and google earth, but all of those web apps not built with all of the security and taxonomy required by enterprise apps.

Our customers want one place to find everything. They want all data sources to talk to each other. Most software vendors in our space are against this and so we have to build it for them. Fred's Wiki is a web 2.0 home brew way of putting all information in one place, but it will not work as a stand alone in more complex industries. He still uses Outlook. Most of our customers are medium size businesses and even though they have an IT department, they do not want to home brew stuff. They want someone to come in and tell them how to do their jobs better. They want to be trained, they want it to just work without too much effort on their part. They want it to integrate with their financial system, their email, calendar, etc...

It just amazes me how Microsoft realized this and basically own the basic OS of the business world (Exchange and Outlook) Really Microsoft's exchange server is the blueprint of Web 2.0. Outlook is "free" much like web apps. Outlook can be customized and modified. You can plug in extensions and use it for all kinds of stuff. Of course you have to buy Exchange, but by then you are too far in to stop. Exchange has a complete ecosystem. They kept the cost low - approx. $167 per user per year (according to Microsoft's website).

My point is all of the hype about web 2.0 is old news. It was all done long ago. Nothing is new. The way to win is to solve problems for customers. To make it easier and cheaper to do something important for a business preferably by a factor of 10. Starting to feel a lot like 1999 again.

Posted by: Dan Cornish | Oct 5, 2005 11:16:39 PM

Seems everyone forgot the word 'pervasive'.

Point2Point or End2End what is going to make this world rock is a pervasive ecosystem - that IS something Google could be bringing us closer to realising.

Of note the stitch-it-together strategy contiues with the bigger players as AOL buys Weblogs-inc.

Certainly enterprise IS moving to Saas (Software as a Service) but perhaps in line with my note on pervasive technologies is Solutions as a Service whereby the industry stops throwing jargon in the form of acronyms and incrementals and gets on innovating and delivering solutions to the problems of today that the ordinary Joe thinks are cool and don't hurt his brain too much.

With a pervasive strategy linked to identity management one can begin to tame the complexity of multiple devices and endpoints.

Just my 2.5c ;-)

Posted by: Ed Danuiel | Oct 6, 2005 12:23:10 AM

The post of SocialText about Wikipedia that you mention is for me another incarnation - hope doesn't sound too religious in english... - of the Long Tail!

The Wikipedia experience shows - for the first time in the non-commercial world ? (the Long Tail example come mostly as of now from the e-commerce: Amazon, Netflix, etc...) - how many small (insignificant on a global scale) social interactions tend to build a huge very significant thing at the global level.

Isn't that clearly Long Tail in action?

That brings me back to the ground topic of your post: future winners in Web 2.0 may be those that can create tools & services where lots of globally insignificant but locally (in geography, community, time, etc.. dimensions) meaningful & useful interactions tend to create a globally very relevant solutions in same dimensions (geo, time, community, etc...).

So, I would not really separate what you call point solutions & end-to-end solutions: I think that winners in web2.0 (i.e service) are those that will be able to build global end-to-end solutions leveraging the micro-value delivered in a Long Tail mode by each local contribution done through pointt solutions

So, end-to-end and point solutions are not 2 distinct ways of doing thing: it's probably just 2 facets of future winners in the web 2.0 world. Btw, it may then probably have become 3.0 when this is achieved....

Posted by: d.durand | Oct 6, 2005 1:58:58 AM

I'm a web entrepeneur here in Brazil. And I feel a bit one step behind you, guys. I feel web 2.0 growing around here, but it comes slowly and late. Anyway, I hope you're right. Web 2.0 seems a lot more interesting than what we have now!

Posted by: Angelo | Oct 6, 2005 12:51:12 PM

dude, my vote is with 'highly integrated but separate services that build and hold emotional connections with their key users'

what my visits to aol, yahoo, msn, google always show me is that one provider can't give it all.

as scott, above, says - we need to make the 'aggregation' of the stuff very easy.

the e2e sites are only 'pre-agregated' but there is never really any e2e integration. that's now how the e2e sites grew and that's not how the bigger e2e sites function as a corporation.

Posted by: charlie | Oct 8, 2005 12:36:25 AM

A VC

Musings of a VC in NYC