just beyond the bend

Web Analytics and Monetizing the Long-Tail of Content Creators / Bloggers

2011-03-14T19:32:43Z

In the wake of the awfulness that happened in Japan this week, I found the traditional news sources (CNN, MSNBC, etc.) to be: A) pretty scant on detail, B) markedly behind the ‘real-time’ on current events and C) ill-informed on many topics concerning the earthquake and related events.

I have thought a lot recently about blogging and web analytics, and monetization of both in the context of “making money off the ‘long tail’”.

I see blogging as being the long-tail of the news and information pipeline, where outfits that monetize through online advertising serve as the short-tail, that which [hopefully] has enough traffic to sustain itself independently.

I see the long-tail of web analytics being those online publications that are unable to justify the cost of services like Chartbeat or GoSquared because they’d lose money in net, or they simply aren’t interested in monetizing their publications. Nevertheless, they may still wish to peer into the analytics and grow traffic however they can.

So why do I reference the content short-tail (i.e., the aforementioned CNN’s and MSNBC’s)?

Well within the same Web Analytics network (i.e., Chartbeat or GoSquared), I believe the long-tail can help the short-tail drive new traffic, thereby adding net readers/ pageviews and perhaps helping justify Free web analytics functionality for the long-tail.

Here’s How:

Web Analytics folks know what’s trending for big publishers. They also have the ability to find experts and influencers from within their long-tail. Imagine if these two folks who are experts [or at least very knowledgeable] in nuclear energy topics and Japanese Disaster Prevention protocol, respectively, could have had their content called up in an instant to fill in gaps where mainstream media channels don’t have access or expertise. This would serve everyone in the content value chain:

Short-tailers: Serve relevant, timely news and information; Drive new and repeat traffic with this high-quality content
Long-tailers / Bloggers: Exposure, Grow reader-base; Rev share on traffic this short-tail content generated
Web Analytics Providers: make paying customers (short-tail) happy with new traffic; make long-tail happy with a free analytics service and new traffic through added distributionà thereby monetizing the long-tail...with short-tail dollars

By using their own tools to pair the right long-tail content for short-tail providers, Web Analytics folks serve to make good money on the additional pageviews they help generate. (And please share a little with the bloggers…)

While the example I give above relates to a singular and rapidly-changing event, I believe the concept is extensible to just about any topic. And while delivery of the content to short-tailers may need to be of a ‘real-time’ nature, the content itself doesn’t require it so (e.g., from above, nuclear theories or Disaster Prevention protocol don’t change overnight).

The Language of Search is Changing; But There’s a Fix! – Hunch, Quora (“social search”), Google

2011-01-18T04:09:10Z

I’ve had two technology topics on my mind a fair amount recently.

Google Search is getting weak and spammy (i.e., less relevant), and
Hunch.com isn’t everywhere on the internet (yet).

Starting with Google Search, I’ve noticed recently that what I search for using search engines and the language (i.e., how) I search with is changing. I’ve read other grumblings about Search from influencers in the Tech arena, as well, and it appears a systematic lag in the algorithm may be developing. I attribute a fair amount of this to social networks and new upstarts that are changing the way we exchange information (e.g., Twitter, Quora, Facebook, etc.). In these information outlets, the phraseology (yes it’s a word) individuals use is markedly different from more formal prose, and so misunderstandings in language can present a barrier to relevant search results.

Further obfuscating a searcher’s intent is SEO and SEM, a compounding effect that has rendered many searches nearly useless for me (I recently clicked on a virus-laden link which was Google’s THIRD SEARCH RESULT DOWN – big thanks to McAfee for blocking the 4 Trojans).

The other thing I’ve noticed is that one of the more promising technology platforms out there, Hunch, is not as widely distributed as the value it delivers would suggest it deserves. Co-Founder and CEO, Chris Dixon, has found that a significant reason for this is that many retailers and other sites that can benefit from a recommendation platform are in fact building out their own engines, as they see the technology as value-enough-to-own.

I agree it’s tremendously important and suggest that the technology could be so ubiquitously valuable, even, that the engine could in fact help significantly improve the Search algorithm layer that identifies context – thereby improving the relevance of a given search. I include a graphic which I will elaborate on below:

In concept, Hunch would sit between two parts of the above system, between the Language & Profile layer (represented by light blue circles above) and two Product layers (Advertising and Search, represented by dark blue rounded-squares). And we’ll use Google as the example search engine (since it’s the most widely discussed as ‘flawed’).

The Transaction bucket would incorporate any purchase information into the broader profile, and include information gathered from Google Market, Checkout, and other payment products the company is expected to launch in the future. Any like / dislike or starring / rating information gathered around Hotpot would be included in this bucket. The Transaction information would then be included into the general Taste Profile bucket.

The Taste Profile bucket houses all the data that Hunch is able to acquire via its existing platform and updates the taste attributes that Hunch processes via Google’s other layers (i.e., Transaction, Language, Search, Advertising).

The Language bucket performs 2 functions.

It crawls the social web to better understand common or colloquial language used in various settings. It observes one-way, two-way, and group exchanges in order to improve comprehension per individual and by the general audience
It also presents a Context layer that, in concept, generates a ’20 Questions’ type analysis behind every search performed in order to develop, over time, a better understanding of the various search ‘types’ a user inputs (i.e., social search, question-looking-for-answer, hash search, etc.). In effect, liken this to the slash function on Blekko search, but the Hunch algorithm would determine this automatically and on a number of vectors, in the background, on every search – based on the specific searcher

The Hunch layer performs a considerable volume of computations in order to blend Language, Transaction, Search intent, and Advertisement interaction in order to build a robust Taste profile that continues to improve with each data point. I won’t pretend to understand the underlying Logic required to support these computational demands, but I would expect this to be very closely weaved into Google’s existing Intelligence layers.

The user-facing Advertising and Search buckets remain largely unchanged, but the Hunch and language layers sit beneath, invisible to the user. Effectively, this adds more computational demand on the backend, but the profile would sit ready to serve search results and Ads based on these on-going improvements to the users search and purchase intent attributes.

I’m not sure what impact this might have on the speed of a search query, but it possibly presents an opportunity to return results more quickly if the prediction engine becomes strong enough to identify an individual’s search patterns. This may not be possible, but doesn’t seem out of the question if Google is able to rule out entire ‘sections’ of the World Wide Web it deems unlikely an individual would require results from [think of this as ‘caching’ buckets of the web, which Hunch could possibly manage by associating ‘sections’ of the web that particular types of people are likely to use].

In summary:

I’ve written previously about folks like Google associating transactions with individuals and their search (and Ad delivery). The transaction piece and tying in Language would provide incredible data to pump back into Hunch’s algorithms to improve the “context engine” and decisioning on Google searches. Thus, improving not only the quality and performance of Advertisements, but also search result personalization.

I believe a variety of social trends will continue to impact the language individuals use while searching and the types of searches from which they expect to get relevant results. Hunch could help bridge the language gap between Search 1.0 (textual) to Search 2.0 (highly contextual) by modernizing Google’s search engine.

Could The Next Threat to Angel and Seed Funding be Kickstarter?

2011-01-12T17:01:58Z

As many people followed over the last few weeks or months, startup Kickstarter.com helped an inventor raise nearly $1M for an iPod accessory – an iPod Touch watch wristband. Hardly a defensible invention; but still an incredible capital-raising story.

Kickstarter is a website that allows individuals to give donations to a variety of business or charity projects - to ‘kickstart’ an initiative. If the total raised reaches a pre-defined goal amount, the money is transacted and the project is ‘on’ (crowd-buysing in deal nature).

In return for the donation, the project creator promises some gift to the backer, often in value commensurate with the donation amount.

While many of the gifts are related to the specific project, what if instead the ‘incentive’ was project equity? What if in exchange for your capital outlay, you became part owner of what you were funding? Not really a novel idea, I’m sure, but one that seems much more achievable and accessible in a Kickstarter Era.

I know handfuls of people with both the desire to raise money for a new business and the desire to invest in great business ideas.

The VC or Angel communities, however, generally demand significantly more money than the average individual can afford [if you can even get access to deals], and alternative investment markets – like SecondMarket – require by law that individuals maintain a net worth of > $2M.

With virtually no barrier to entry on Kickstarter - other than hitting the project pledge goal - if the project owner was able to set up an equity structure that “investors” felt comfortable with, Kickstarter or some like platform could begin to increase the pool of startup companies, or even displace some existing Seed or Angel opportunities.

While I’m not sure the impact to the startup industry would be net positive for Investors [or even Entrepreneurs], it certainly has the potential to materially disrupt early-stage investing, an area many Angels grumble is already over-subscribed.

It’s Open Season on OpenTable – Decentralizing the Platform and Optimizing for the Merchant

2010-12-15T13:58:21Z

While investors continue to swarm on OpenTable (now trading at a P/E >130), I’m not convinced that they have a business sustainable beyond even the next 2-3 years.

I don’t believe OpenTable’s demise will come from a direct competitor – of which they’ve had effectively ‘zero’ during their rise to greatness – but from every other angle imaginable. My feeling is that the entire weight of the growing local online advertising and Deals industry is going to fall hard and fast on OpenTable, which is already rife with customer complaints like this now popular one.

The chief grievances from many restaurateurs are, in order:

High fees for filling tables (majority of profit per head)
Feeling locked in once part of the system

The cost of OpenTable is nearly prohibitive for many restaurants when they soon discover extra tables don’t translate to new loyalty. And by this time, many owners already believe it’s too late to cut the cord.

So where does that leave us?

Well, with an open door to decentralize the system, give control back to the restaurants, and perhaps support the industry with a more proportionally fair volume of the profits generated from directing business to the restaurant. An open door that I believe is a more sustainable way to support the online reservation industry and its clients.

And who would jump at this opportunity?

Advertising giants like Google and Deals mavens like Groupon present compelling cases. Both companies are known data collectors and synthesizers, continually aggregating, cataloguing and optimizing. Either one of these companies could build the necessary tools into their existing merchant pages [no easy task, but certainly possible given their cash positions to buy talent or technology] and create a way for restaurants to manage their reservations within the same environment where their web profiles are managed.

In the case of Google, they could give restaurants some power to ‘auction off’ tables – i.e., bid in real-time for Ad space on Google.com or Mobile – in order to fill last-minute tables for Google users likely to enjoy the restaurant. An inventory management technique used in many ticketing verticals from live music to the airline industry.

For Groupon, given the extended redemption periods for most / all of their coupons, they would have an opportunity to identify longer-term trends and suggest offers to merchants as a way to stimulate business during historically slow periods or seasons. Ideally, Groupon would also use the data collected from each reservation or page visit to build a taste graph around deals consumers would be most likely to enjoy (this could even vastly improve their display ad spend by enabling personalized offers via retargeting).

And for consumers, perhaps this decentralized platform could enable a new bidding system in itself where patrons interact with the restaurant layout and vie for the best table in the house.

Ideas abound, OpenTable is very exposed in a decentralized environment. While the company now supports the ‘OpenTable Connect’ platform, which eliminates the need for proprietary hardware, as a public company relatively low on cash, it is significantly handicapped in its ability to shift from a fee-based system to a decentralized or ‘other-monetized’ platform. For Google or Groupon, online reservations would only be a 2nd or 3rd or Xth core business, but one that could become a huge data generator and overall boon for personalized and cross-platform Ad optimization.

I hope Google or Groupon joins the online reservation business. It’d be good for consumers and great for restaurants.

‘App-etizing’ Websites – Interactive Widgets that Can Drive New Revenue Streams for Publishers

2010-12-06T23:48:41Z

A really exciting concept I‘m hoping has legs is the notion of hosted applications on publisher websites. A couple different examples today are HeyZap and PostUp, two products that help add interactivity and incremental revenue streams to publisher pages through gaming and social features, respectively.

In the case of HeyZap, the company allows visitors of a publisher site to link to their Facebook games via Facebook Connect and play them directly on a publisher’s website, bypassing Facebook.com and thereby extending the reach of their games across the web (big plus for app developers!).

HeyZap and PostUp are just two early examples of what could be a much larger business opportunity: creating hosted environments where a widget can be dedicated to anything the publisher sees fit.

If the website is a gaming or gaming news site, it is possible the publisher can allow games to be played through the widget. If the site happens to be music-oriented, it’s possible the interactive real estate could enable a tool that allows readers to view local concerts and purchase tickets directly through the small widget. Or navigate and purchase / gift related songs via an iTunes Store widget. One could get very creative on what functionality sits within the widget!

What’s very ‘open’ about this idea is the ability for any developer to create a widget. Consider just 3 possible scenarios:

Publisher creates and features their own App (which is hosted by a 3^rd party service like an App Store)
Publisher selects the 3^rd-party Apps they will allow to be featured on their site (similar to picking Apps from an App Store). They can drive new revenue and ensure content is ‘brand-safe’ for their site
Website visitors select the set of Apps they are most interested in interacting with (and perhaps the most likely to drive revenue for publishers)

There are a variety of possible business models associated, including: revenue-shares, CPM, subscription, and virtual exchanges.

Some reasons I believe this model has legs across the entire value chain:

Publishers help distribute content via widgets, creating for Consumers a very connected web experience across multiple sites.
Developers [both independent and affiliated] have another distribution environment (beyond their own URL’s and mobile Apps)
Publishers benefit by unlocking a new revenue stream on their site, one which they have a tremendous amount of latitude to control.
For the Service hosting the widget or application environment, they get a cut of the revenue generated for delivering the technology layer that enables App development and distribution, as well as transaction or exchange functionality.

I’m really excited about this market and believe there’s a tremendous volume of innovation this industry could herald in.

National Geographic Has Limitless Opportunity to Monetize 100+ Years of Valuable Content

2010-11-24T14:58:18Z

I can’t think of another media company today that has a larger – and perhaps more valuable – archive of content than the National Geographic Society (NGS). NGS has been producing magazines since the 1800’s and has dipped majorly into many forms of artistic content, including photography and many forms of video.

NGS is one of the world’s most recognized brands and has been very successful “inspiring people to care about the planet since 1888” – as the motto states. It’s also a non-profit that promotes global exploration, conservation, and awareness – 3 great platform principles anyone can get behind.

While it would be unfair to compare revenues with other media companies, since business and credo differ handily from public corporations, I think a qualitative statement on relative exposure is fair: National Geographic content is rarely seen in the wild outside of major releases like the ‘Earth’ documentary. And these videos make up just a very small portion of the company’s total archives. Their collection of content is larger than any other media outfit out there, but they choose to promote only very recently produced material [discounting the 100+ years of exciting material they’ve amassed].

There exists a big opportunity for NGS to leverage much of this dormant archive in an area I believe they have largely ignored: Exploration. The Travel space is a natural fit for NGS, but the company has surprisingly little influence despite their emphasis on Global topics. While much of its content is old [by date], it is not so much outdated. Historical and cultural information is timeless!

As an avid traveler, where I‘d love to see NGS leverage this largely dormant content library is in helping educate travelers about a destination or adventure before they step on the plane.

It would be exciting to do a search for all the content related to, say, Etosha Safari, and have returned all the articles ever written about the Park, Namibian culture, nearby destinations of intrigue, politics of region, etc. Likely, there are photos NGS has associated with many of the articles, and even videos or documentaries on related local topics.

Perhaps create an App store-like marketplace where travelers could buy articles, pics, or videos in bundles (say $10 for 3 articles and 1 video), or just purchase individual archives. Partner with travel sites so this could be offered through Kayak or Orbitz, or other sites NGS believes appropriately represent its brand.

I find I often arrive somewhere abroad without something beyond a Wikipedia education on where I’m going. It would greatly improve the travel experience with the collective archives of NGS at my fingertips. I could create my own personal Etosha cultural packet and have the most in-depth and interactive content available.

But this idea is part of a larger point: National Geographic has a library of content waiting to be unlocked!

Check-ins without transactions can’t support location-centric social services

2010-11-18T23:37:15Z

Check-in and reward sites are popping up like clockwork these days, most prominently in mobile and ‘cross-web’ form-factor. A few examples below:

Physical locations – Foursquare, Facebook, Shopkick, Yelp, Gowalla
Websites – OneTrueFan, Meebo

While I would argue that checking-in is the same thing as ‘Search History’, and any browser owner that wants to add game mechanics has an easy switch to turn, the check-in ‘Industry’ is still very nascent and divided, and winner(s) are very much undetermined.

Nevertheless, it’s hard for me to believe that unless transactions directly accompany check-ins – or transactions are the check-ins – that location data alone will build a ‘taste graph’ strong enough to support the volume of investment we’ll likely see in this space.

There’s a mass migration around promotions, discounts, and recommendations in the space, and there’s definitely strong evidence that these will be the dominant business model emphases for some time. How these features will be delivered will vary.

Much of this current delivery is via ‘check-ins’, and relies on game mechanics and badges. Far less appears to be invested in how monetization and ROI will come from synthesis of acquired and meaningful data.

On a questions-for-recs site like Hunch, or with a purchase through Amazon, the participant or purchaser is consciously making a decision and an indication about preference. This is meaningful! In a badge-driven world, the data can be handily skewed from actual preference of brand or likeness, and instead toward dishonest game-play (oh so very often!). I see many false Foursquarers simply checking-in to get badges. This won’t help the company build customer loyalties to particular stores or help local advertisers, but rather just present huge brand-names a new marketing channel. This is the prevailing model today.

But the game will get old and in time the Marketing investments will begin to look like ‘One-offs’, and the ROI’s less attractive.

As merit badges begin to lose their luster, it will become increasingly important to match transactional data with check-ins, or the taste graph will not be fine-tuned enough for Normals to find these services appealing. [And surely early tech-adopters won’t be enough people to support the investment this space is welcoming].

Normals will demand accuracy and relevance of promotions once the game is over!*

So that ‘taste graph’ built to deliver those promotions, discounts, and recommendations better be great, as the check-in business will undoubtedly be using this transaction data to support the business.

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

*History tells us Normals have a low tolerance for annoyances: do-not-call lists, Spam folders, email opt-outs. Even the US Postal Service doesn’t include spam / flyers if you ever hold or forward your mail!

Hunch’s Two Hurdles: Consumer Participation and Site Differentiation

2010-11-08T23:08:27Z

Hunch is a really cool tool for consumers and brands, alike. The application helps people “personalize the internet” by making intelligent recommendations for various things their algorithm believes you’d like. Answer some questions in their application and the recommendations get better. The more questions you answer, the more Hunch learns about your personal ‘taste’. They serve up recommendations for all topics from ‘which camera should I buy?’ to ‘what political candidate do I most closely align with?’ and ‘what day-trip would I most enjoy this weekend?’ It’s a really neat application that a lot of people find to be amazingly practical and accurate, even for really far-out questions.

Online retailers would naturally love this, because their customers could quickly figure out which of the 20 cameras they supply would best fit a given shopper’s needs. Happy customers means transactions and return visits.

Despite this seeming “Match made in Heaven”, I believe there are 2 hurdles Hunch needs to bound before they will see mainstream adoption:

1. How Do You Get Consumers On Board?

While ‘incentive’ has become somewhat of an ugly word to me given the recent explosion of group-sale and check-in deals, I believe Hunch will need to incentivize participation in order to get consumers to take that initial step to answer questions. This may also demand stepping back from recent white-label solutions they offer partners.

It will take an added push to get shoppers ‘over the hump’ to use the tool, particularly given that it takes several minutes to get results from answering questions. I also believe it will be important for consumers to see the Hunch logo on the application across a variety of sites, so they can identify and trust the experiences they have with this question-asking application (Hunch) will be a positive use of their time.

The nature of a Hunch incentive may only need to be a one-time promotion if the company believes they can build this trust in a single interaction, but it is more likely that they will need to present on-going incentives, and possibly tied to specific retail sites or purchases.

2. Differentiation, Data Ownership – And Discovery:

A second longer-term issue I see is one of differentiation [of data, not competitive]. Since most major online retailers carry the same Top 10 or 20 products (e.g., TV’s available on Best Buy or Amazon or Walmart.com), what’s to differentiate one site using the Hunch application from another? Hunch currently owns all the data, so most of the recommendations will likely be the same across major retailers, given they have similar combinations of the top-selling products.

While I’m sure this differentiation issue may be a bit of hyperbole, I’m not sure retailers would disagree that this is a much bigger problem, particularly given the data Hunch users yield strengthens Hunch’s recommendations for all partner sites (including likely competitors). This could cause some adoption [and data-ownership] friction!

Where I think Hunch has a really great blue-ocean opportunity is helping corral true “long-tail” sites where consumers are overwhelmed with options and even a starting point for a product search is hard for a consumer to ascertain. My gut tells me that Hunch would have tremendous success making recommendations in places where consumers confront literally thousands of seemingly ‘like’ options – where true Discovery can be a high-value tool / functionality.

Examples are:

Apps – iTunes, Android Market
Songs – Pandora, Spotify
Niche Retail – Etsy
Social – Twitter, Quora [these are non-transactional, but would still help strengthen the algorithm]
Ads – Google AdWords (in gmail, search)

These are ecosystems where I would focus my technology and sales efforts in order to build a long-term, sustainable product that provides unique – differentiated – benefit for Hunch’s paying clients.

An Idea for Keeping Engaged Users [Versus Just Keeping Users Engaged] – Disqus / Outbrain

2010-11-04T02:52:57Z

A regular top-of-mind concern many news and magazine publishers identify is not just driving site traffic, but driving ‘quality’ traffic that doesn’t dilute the value of a page-view.

Over the last couple of years Comment platforms have become increasingly powerful, enabling publishers and users to easily: organize, share, create streams, socialize, like, and manage their comments.

While I can only state so qualitatively, it appears the quality of commenting and the surrounding discussions have improved significantly. I often find more interest in the conversations at the bottom of the page than I do in a given article. It’s also exciting to see when contributors help keep authors straight, or even inspire follow-on posts or articles. One good example is Fred Wilson’s AVC.com blog, which added topics like ‘MBA Mondays’ as an additional ‘Theme’ for the site – inspired by comments from the gallery of readers.

These sophisticated commenting platforms are helping build this ‘quality’ traffic publishers are looking for.

My favorite among them – and, in fact, my most often recognized in the wild – is DISQUS. The platform includes many / all of the aforementioned features, and even contains analytics and other operational bonuses for paying clients.

On the DISQUS sites I normally read – AVC.com, TechCrunch to name a couple – I see many of the same faces across these sites. These seem to be the most loyal readers – those engaged and always contributing.

These individuals are the most important visitors on the site since they are not only highly engaged by the site’s specific content, but because they also create new content. For Free. That other people may enjoy or engage with (or even just feel engaged with). Getting these contributors to create more is a boon for all: You the publisher, me the contributor, and others interested in finding your contributor’s comments. A publisher’s goal should be to keep these high-value folks on the site longer!

It would be exciting to encourage this through a sharing of data between a company like DISQUS with another that makes intelligent recommendations for ‘Articles you might like’ – a company like Outbrain.

Why?

Because together they would Keep Engaged Users versus just Keeping Users Engaged

How could DISQUS / Outbrain do this? At least three ways:

Present article recommendations by pairing the current article’s keyword attributes with other articles in recent press on the site [already done, I hope]
Present article recommendations based on keyword tags from other articles a commenter contributor to
Present article recommendations based on the attributes of another commenter a user has engaged with / replied to in the past

Keeping existing customers happy is important, as acquisition is a lot more expensive. The contributions these high-value commenters make may even lower the cost of customer acquisition as they continue to produce new, quality content for a publisher.

This pairing of data between commenting platforms and article recommendation would help improve the ‘quality’ of a publisher’s traffic and likely generate new content other readers may find ‘sticky’.

Nexus One – Google’s Marketing Campaign that set the standard

2010-11-01T21:03:46Z

The Nexus One had to be one of the best marketing campaigns of all time.

What it did for Google:

Introduced the Android OS with Google-size press coverage
Allowed Google to set the base standard for how any Android device should look, feel, perform
Paid for itself – Although it had a short shelf-life, the phone was actually profitable!
Helped propel it to the #1 smartphone OS in the world

What is does for us:

Provides benefits of healthy competition: sub-$100 touch-screen smartphones, presents alternative to iPhone for consumer smartphone OS (Palm and BlackBerry primarily only Biz focused, designed)
Drives innovation in the consumer smartphone space: hardware , software / UI, App design, et. al
Ubiquitous across all Operators (not just AT&T)

I’m sure the Google’s purported Nexus Two will set a new design and performance standard for both hardware and the softer components for future device releases. And I’m looking forward!