from citi analyist about blinkx rated buy
Contents
Blinkx: A Mix Of Autonomy & Google Or Pie In The Sky?
4
What Is Blinkx?
11
Risks – A Hill Of Beans
15
Internet Video – Where Is Consumption Trending?
20
Advertising & Video Search – What Is The Potential?
28
Technology And Patents
34
Broadband TV – Birth Of New Aggregators
46
Transaction Hijacking – E-Commerce Opportunity
48
Deal Structure And Parent Company Relationship
51
Financial Model
54
Valuation – Early Stage Business
63
Appendix 1 – Search Trial Results
70
Appendix 2: Patents
73
Appendix A-1
77
Blinkx Plc (BLNX.L)
11 July 2007
Citigroup Global Markets | Equity Research
4
Blinkx: A Mix of Autonomy & Google or Pie in The Sky?
We pose two questions. Does Blinkx business model have differentiation in
technology, functionality and consumer appeal for the group to attain a large
piece of the video search market? Are video search revenues likely to scale or
not? Autonomy’s technology and Blinkx consumer application should give
Blinkx the chance to compete with the likes of Google. As to the end market,
video search’s growing share of media consumption and the advertiser appeal of
emotive television content and contextual search should produce scale.
Ultimately, Blinkx is either worth very little or way more than $236m to a large
internet group. If our model is right, Blinkx trades at 7.9x P/E in March 2011E.
The summary in summary
Routemap to this section
Blinkx business model – video search & consumer website
Technology drawn from Autonomy
Scalable end markets: video search could replicate textual search
Emotive TV and contextualized search; could yield rich valuation
White label: strong momentum since the IPO
Issues: competition, content, consumer behaviour and monetization
Deal structure, relationship with Autonomy and management
Earnings model, sensitivities and implications for valuation
Blinkx business model – video search & consumer website
The basis of the Blinkx offering is 12 million hours (and growing rapidly) of
indexed online video content. Blinkx has two main businesses: provision of
video search technology to third parties, and a direct-to-consumer website that
offers streaming video search and acts as a video portal organized into subject
categories: world news, sport, business, health, entertainment, user generated.
At present, Blinkx generates around 4m video searches per day. Further
applications such as Broadband TV and e-commerce transaction hijacking (see
below) will follow in the next year.
One such application is ‘white label’ agreements, which allow third parties,
predominantly portals like MSN, to licence their video search technology. In
March 2007, these represented around 65% of the group’s video searches.
We would expect ‘white label’ to remain the main revenue source for the
group for some time. The distribution partners benefit from being able to offer
advanced video search on their websites, while Blinkx benefits from
increased usage of the search engine and therefore increased potential
advertising revenues. More recent deals have seen Blinkx also receiving a fee
per search as well as its share of advertising revenues. Since the IPO, Blinkx
has concluded / extended a number of important ‘white label’ deals. The
most important have been with Ask and RealPlayer, as we discuss below.
Blinkx Plc (BLNX.L)
11 July 2007
Citigroup Global Markets | Equity Research
5
Blinkx.tv or Blinkx.com is a consumer facing website, which allows
consumers to search for internet video content. Blinkx has more than 130
content partners, including traditional media such as ABC, HBO, ITN, BBC
and Reuters, and new media such as YouTube. As consumers type in a
search, such as Wimbledon, a wall of 25 mini screens of streaming content
populate results of each search. The consumer can then click into one of the
boxes (which then occupies half the screen). A Blinkx search is based on its
entire indexed content, irrespective of whether said content emanates from a
content partner, and the search result is biased purely to efficacy of results,
not to agreements. The site also acts as a portal allowing consumers to keep
up with events in areas like business, sport and world news.
Later this year, the group will roll out two new services: Broadband TV and
Transaction Hijacking. Broadband TV is an offering of primarily basic
channels (infinite in number and high definition) streamed using second-
generation p2p technology. The service will initially be ad-funded, and the
flexibility to move from streamed TV into advertiser product websites at the
click of a mouse should prove attractive. Subscription offerings will follow,
and we think technology solutions allowing PC to TV convergence will prove
the driver for this in 2008/9E and beyond. Blinkx is a European play on
internet search advertising, an attractive growth market and difficult to invest
in outside the US.
The group will also offer ‘transaction hijacking’, which uses implicit query
technology to inform consumers of cheaper transactions at the point of
purchase. The model intends that Blinkx will be paid a fee by the site which
captures the business.
We expect further mobile, desktop and local search applications to follow in
the future.
Technology comes from Autonomy, its parent
Autonomy’s matching algorithms have for over a decade allowed it to categorize,
hyperlink and route data that has not previously been tagged / analysed. The
algorithms can also profile users’ areas of interest based on how they engage
with data. The technology is proprietary and unique. Other technologies which
usually search for simple key words are not nearly as effective, particularly in
dealing with multi-media content. Autonomy’s end markets are predominantly
software applications within enterprise and government. Blinkx focuses on
consumer markets and has a perpetual licence for this technology globally,
except in the Chinese language, and this is exclusive for five years. Autonomy is
a 10% shareholder in Blinkx and its CEO and founder, Mike Lynch, also owns
10% of Blinkx.
From 1999-2002, Autonomy traded at anything between 30x and 90x
prospective revenues. Currently, the group trades at 29.7x P/E and 6.3 EV/Sales
2008E, and in the interim phase its market cap has moved from c.$1.5bn to
$2.5bn (with massive volatility along the way). Our price target for Blinkx is 80p,
which corresponds to an EV/revenue multiple of 93.3x March 2008E, falling to
3.6x March 2011E, or 12.2x 2011E EBITDA.
Blinkx Plc (BLNX.L)
11 July 2007
Citigroup Global Markets | Equity Research
6
Scalable end markets: video search could replicate textual search
The online video market has seen significant growth fuelled by demand from
consumers, advertisers and website publishers. The key issue is to what extent
this market can be monetized and be a catalyst for shifting advertising budgets
onto the web. According to eMarketer, online video advertising grew over 80% in
2006. There have been several drivers:
In 2002 broadband users in the US numbered 60m. Currently, there are
130m, and by the end of the decade this number is expected to grow to
180m. Roughly 80% of US broadband users stream video content online, and
this percentage is growing. The top five video sites saw average time spent
grow by 5% in the five months between June and November, so individual
usage is also growing.
One of the major drivers of this growth has been better user-generated
content, and recently a spate of ‘professional’ content companies including
NBC, Newscorp, Disney and CBS have all begun to embrace the internet.
Further technology improvements, such as improved compression (to
facilitate better-quality video) and home networking (to allow PC content on
TVs) should be a further driver of growth in this market for the next five years.
Blinkx is well placed for these developments, especially as PC-TV
convergence becomes widespread.
This kind of growth is reminiscent of textual search and Google. Textual search
revenues in the US in 2000 were $81m, and by 2006 were $6.7bn. eMarketer
predicts online video advertising growing from $85m in 2003 to $2.9bn in
2010E. There is an enormous range of forecasts in the market – eMarketer
raised its 2010E forecast by c.25% between November 2005 and November
2006, while IDC’s March 2006 forecast has the market at $1bn in 2010E. By
2010E, using eMarketer’s estimates online video will only account for 3% of
total US TV advertising and 12% of US internet advertising.
Google’s revenues have grown from $440m in 2002, to $10.6bn in 2006, and a
limited part of this growth is acquisitions. Google’s multiples have borne some
resemblance to Autonomy’s. In 2002, average EV/revenues was 69x prospective;
in 2003, it was 35x; and, currently, Google’s P/E is around 33x (an EV revenue
multiple of just under 7x).
Blinkx could attract an extremely high multiple for two main reasons:
Video search brings together the emotive appeal of brand advertising
reminiscent of TV, with the contextual, measurable quality of textual search.
Blinkx’s technology affords multiple applications for monetization, such as
‘white labeling’, a direct consumer website proposition, Broadband TV,
ecommerce transaction hijacking and mobile applications. News-flow around
white label deals since the IPO has been very encouraging.
Blinkx Plc (BLNX.L)
11 July 2007
Citigroup Global Markets | Equity Research
7
Strong momentum in white label deals
Since the IPO there have been a number of important white label deals.
On 5
th
June, Blinkx announced that it has been chosen by Ask.com to power
the video search component of its ‘universal search’. When consumers search
on Ask, which is the fourth largest global search group (with a c. 5% share in
US), behind Google, Yahoo and MSN, the results come back as text, image and
video searches. This ‘universal search’ affords Blinkx a great deal of video
search traffic, but less ‘monetization’, as consumers will frequently search on
text or images rather than video. We see Ask’s decision to use Blinkx to power
its video search as a huge endorsement, given the importance of video search to
Ask’s new offering. At present, we assume that it has a material affect on
search traffic, but not on revenues for the reason we give above, that click
through rates are likely to start very low. Over time we would expect these to
pick up. Ask recently spent $100m on a marketing campaign in London, San
Francisco and New York promoting its new search service.
On 25 June, just as important as regards traffic, Real Networks announced that
it will have video search integral to it, and this will be powered by Blinkx. 1.5m
Real Players get downloaded every day, and the tag 'powered by Blinkx' will be
front and centre on the video search function, which will also be front and
centre. This should be a major traffic driver and driver of brand.
Both deals work in the same way as traditional white label – the traffic is Blinkx,
and ads pre rolls are shared between Blinkx and the content owner. In the case
of Real Player (and Infospace – see below) Blinkx also gets a share of adverts on
the sites split 50:50 with white label even if no advert is contained in the video
search. In the case of the Ask deal, Blinkx gets a fee per search as well.
On the same day, Blinkx also announced that its European deal with Infospace
has been extended so that Infospace has opened its US traffic to Blinkx engine.
We would not expect this to drive massive extra traffic, but again, we see it as
vindication of the product. We note that Infospace only allows Blinkx on its own
traffic, not the networks it powers, but overtime, this could change.
The Independent newspaper also reported on 25 June, that Blinkx is
conducting 4m searches per day, up from 1.4m at the time of IPO. We had
initially modelled a doubling for the full year, so this is very significant, as the
Real and Infospace deals should drive traffic still further. To the point we made
above, however, we do not feel comfortable yet adjusting our forecasts, because
much of the extra traffic will have come from Ask, where click through rates will
be lower.
As regards monetization of traffic, Blinkx also issued a release about an ad
service it provides, called Ad Hoc, which adds relevancy to ads. This will start
having some impact in Q4 of this year.
