Jan 182013
 

A recent article on Techcrunch made the argument that the term “big data” should be abandoned, on the basis that it has fallen into generic usage and, in any event, no longer represented how organisations are thinking about their data.

Organisations aren’t interested in amassing lots of data, the author – Leena Rao – opined. Rather, they want to derive insights from big data.

She has a good point.

There is a real risk that the "big data" movement will soon experience the same backlash that befell the CRM movement in the 90s.

The key problem is that too many people think of "big data" in terms of a technology purchase. The mind set is: "We will choose a big data vendor, install their solution and then – automagically – we will reap significant business benefit and advantage".

I always prefer to describe CRM as a "philosophy". Organisations need to embrace the key tenets of customer relationship management, which in turn requires making changes at multiple levels – technology (of course), employee training, internal policies and procedures, customer communications and, importantly, culture.

That is, to successfully embrace CRM, an organisation must ensure its day-to-day business operations and company culture fully reflect the tenets of CRM.

As such, it is a philosophy – a way of thinking/acting – as much as it is a technology solution.
Today, many organisational stakeholders are falling into the same "buy the technology, secure the prize" trap when it comes to big data.

Big data is actually worse, because it requires multiple kinds of technology purchases: solutions to handle data ingestion, validation, storage, management, aggregation, chunking, abstraction, querying, hypothesis testing, access controls, reporting, visualisation etc.

It also requires significantly wider changes to business practices, policies and procedures – often involving multiple external partners within a supply chain – to achieve even the slightest outcome.

Jan 032013
 

(I wrote the following blog post for my new employer’s blog – see the original at http://www.compassitesinc.com/blogs/breaking-free-of-the-website-product-mindset-using-real-time-data-and-search/)

In dozens of large-scale website development projects, I have witnessed well-intentioned project teams repeat the same mistake in prioritizing their time, focus and investments.

The primary driver of project mistakes is the incorrect assumption that the finished solution is a deftly structured content product defined – and organised – through comprehensive and internally coherent information architecture (IA).

With this assumption in mind, and correctly recognising their target users will have differing content and task needs and preferences, they prepare user personas and the like, and then map distinct user journeys, in order to identify and test key touch points that create opportunities to meet and exceed user expectations.

Almost as an afterthought, they then plan the implementation of a search tool. In many projects, search is seen as a fail-safe, user-driven discovery mechanism should the IA and carefully crafted user journeys be unsuited to a given user’s task or objectives.

As a consequence of this approach, the bulk of project preparation and planning is spent ensuring excellent content quality and coverage to meet user needs, as captured in the personas. The majority of the remaining time is spent refining and testing the IA.

Only as launch nears is some time invested ensuring search is functional and results pages offer an efficient and intuitive discovery context.

New understanding of Cognition

Recent research has confirmed what many have long believed. Individuals have very distinct cognitive styles, which dramatically alters their preferences for digital content presentation and consumption.

Individuals can fall into the following categories (which are by no means exhaustive):

  • Impulsive v. Deliberative – Impulsive individuals prefer to make quick decisions, and don’t wish to be over-burdened with information – they just want access to the information they consider key to their decision. Deliberative individuals, on the other hand, prefer a more measured approach to decision making, and want to explore options in detail before making a decision.
  • Visual v. Verbal – Visual individuals prefer imagery over reams of text, and respond well to strong, emotive images, whereas verbal individuals prefer chunks of text and numbers – unadorned by troublesome images – so they can quickly decide for themselves (using data) how they feel about the company, product or service.
  • Analytical v. Holistic – Analytic individuals prefer to access the technical details in all their glory, even if they do not have a full understanding of their importance, whereas Holistic individuals prefer to see a concise statement of ‘the bottom line’.

It is important to note that an individual’s cognitive style may change, depending on the task or consumption context. In one consumption context, say buying a new car, the preferred style may be analytical, but in a different context, such as planning a holiday, it might be holistic or even visual.

What replaces the IA?

As our knowledge of cognitive styles increases, it becomes apparent that static IAs and single-form content are insufficient to achieve a truly user-centred, and user-friendly, navigation experience.

Undoubtedly, it is important to ensure that your content is – and will remain – curated to the highest quality and relevancy. However, it is clear that multiple variants of each content item/asset will be required, to meet the specific presentation preferences of different user types.

Static IAs, however, are poorly suited to presenting large content offerings, much less multiple content display options.

The solution is to architect search technologies into the website design from day one.

Search functionality should permeate each and every page of content, though not just as a search bar or other user interaction. Rather, search should – quite literally – be embedded into each and every hyperlink within the site.

A hyperlink should not contain a static pointer to a URL. Instead, it should contain metadata about the target page (i.e. its core content element), which is used to trigger a search each and every time a user clicks on it.

This search activity creates the opportunity to blend both the individual’s expressed intention (i.e. the content item they wish to view, as denoted in the URL’s metadata) with real-time behavioural data generated during the user’s visit (such as the navigation labels, headlines, images and pages the user selected previously).

A more fluid experience

Under these circumstances, it will be possible to discern not only the reason and context for the user’s visit (such as to compare multiple products or make a purchase), but also their cognitive style (by analysing which styles of content presentation led the user to take specific actions within the site).

As a result, it is possible to dynamically alter – and optimise – the basic structure and content of the website.

Irrelevant navigation options (or even entire areas of the site) can be hidden, while more pertinent links can be highlighted or promoted. Content items can dynamically optimised, either by selecting the appropriate content item variant (if in a fixed format) or, alternatively, applying the template most suited to the user’s cognitive style.

Adopting this fluid, search-driven approach to website design and content and service presentation will result in a more compelling and intuitive customer experience, which in turn will improve business outcomes.

Sep 192012
 

An interesting new study, conducted by Duke University on behalf of the American Marketing Association, has been published, and is well worth reading.

It provides a number of insights into factors influencing market development, growth strategies, fluctuations in marketing budgets and growth in social media marketing expenditure.

But the table below provides a real eye-opener for those unconvinced of the emerging influence of design thinking, and its related discipline, service design.

Key takeaway: Traditional advertising expenditure (that is, offline advertising) is expected to fall by 137.5%. Online spend will also fall by 10.2%.

What will Chief Marketing Officers be directing their marketing budgets to instead?

Designing and implementing better customer experiences (+26.8%) and, more tellingly, designing and developing new services (+52.4%).

Jan 132009
 

I had one of those ‘gotcha’ moments today.

I’m in the process of moving and, despite my best efforts, find myself in a broadband ‘dead zone’. Not in the sense that I moved to the boondocks, but due to the time it is taking my provider (iiNet) to commission an ADSL2+ service at my new place.

So this week I find myself sitting in my office surrounded by packing boxes and limping along on my standby 1Gb 3G wireless service (via Three, connected @ 7.2Mbps). Normally I only use this when I find myself travelling and staying at hotels with outrageous broadband charges. Now I am trying to run a business and maintain a connection with the outside world.

And you know what…it is ridiculously difficult to get by on a 1Gb plan over a painfully slow wireless connection.

But that’s not the point of this post. My point is that for all the frustration this situation is causing me, I still have *greater* connectivity that most Australians.

That’s right…the majority of Australian households use Telstra’s cheapest broadband plans, with 200Mb per month quotas and barely-better-than-dialup speeds.

Which brings me to that ‘gotcha’ moment I experienced earlier today: how does this ‘connectivity lag’ impact your business and technology plans, and how will it impact your growth in the local market? If the average Australian household has ‘fraudband’, and your product/service assumes high connectivity, how does this shape your ability to serve the local market? How could you reshape your offering to target a broader section of the local market (at least until the connectivity lag reduces)?

Food for thought.

Nov 052008
 

Given how late in the day it is, I suspect everyone is almost over the 2008 US election.

But I wanted to add a note about my experience in trying to monitor the election online today.

I watched the CNN TV coverage live, but wanted to compare/contrast their stats with that of other major US media outlets. So I monitored the Web sites of CNN, Fox News, BBC News, New York Times and MSNBC.

Each had a very similar approach to covering the vote counting – a map of the USA, with each state coloured according to their current status (i.e. leaning towards the Democrats or Republicans). Most also allowed you to mouse over individual states to get more specific information on the progress and projections for that state.

Now here is what bothered me: the data on all of the US media web sites trailed what was being announced on-air, often by a matter of minutes. The CNN site was particularly annoying, as they had both a ticker and a map on their homepage, and while the ticker kept in close synch with their TV ticker, the map was often 10 minutes behind the ‘live’ data. Same site. Same page. Presumably same data source. Different update speeds.

(The BBC was a potential exception, as their web site was often minutes ahead of the other sites, but I did not have access to their broadcast, so I could not see whether the two channels were in synch.)

How did this happen?

Even the mainstream media has commented quite frequently about the importance and role of the Internet in the 2008 US Presidential election, in terms of both generating donations and motivating volunteers to join campaign activities, and in dramatically increasing voter registrations.

Much was made of the incredible viewing figures that political satire and comedy segments (especially the Palin impersonations on Comedy Central) were pulling online – often far exceeding their broadcast viewer numbers. Social networks like Facebook, and microblog services like Twitter have generated unprecedented traffic around political groups, conversations, debates and commentary.

Just as TV ‘made’ John F Kennedy – his superior telegenic appearance giving him the edge over Richard Nixon in the 1960 debates, the first to be televised – Obama’s campaign was given an immense boost due to the manner in which it was able to engage voters via the Internet (embracing and extending the lessons from Howard Dean’s campaign).

The abovementioned media outlets had plenty of warning that there would be significant demand for online information and analysis during the election and vote counting, but they clearly dropped the ball.

Fastin Phentermine Drug
Phentermine Diet Pills Cheep
Addicition To Phentermine
Phentermines
Foradil Phentermine Norvasc
Phentermine Legal Mexico
Phentermine In Urinalysis
No Rx For Phentermine
B12 Loss Phentermine Vitamin Weight
Phentermine Swelling
Phentermine Drug Interaction
Phentermine No Doctor’s Prescription
Can I Take Phentermine With Tenuate
Phentermine Non Perscription
Phentermine For Sale Without Perscription
Phentermine Cause Birth D
Phentermine Yellow No Rx
Bontril Foradil Phentermine Evista
Phentermine Purchase Without A Prescription
Phentermine In Stock Overnight Delivery
Online Phentermine Adipex Diet Drugs
Cheap Phentermine Purephentermine
Order Medications Free Shipping Phentermine
Phentermine Causes Bad Breath
Sliming Sources Phentermine
Cheapest Phentermine Around Phentermine Online Cheep
Phentermine No Rx Cheap
Phentermine 37.5 Blue White
Phentermine Shipped To Florida Online Pharmacy
Phentermine Low Cost Online
Weight Loss Drug Phentermine
Cheap Brand Name Phentermine Canada
Phentermine Or Meridia
Weight Control Center Phentermine
Rebound Effect Of Phentermine
Phentermine Usa Over Night Without Prescription
Cod Order Pay Phentermine
Delivery Phentermine United Parcel
Drug Phentermine Screen Urine
Phentermine Non-Generic Capsules
Phentermine No Prescription Get It Online
Sie Effect Of Phentermine Hydorchloride
Phentermine No Rx Buy
Phentermine Not For Sale
Phentermine Rating
Drug Phentermine Prescription
37 Phentermine
Phentermine Wsith A Script
Cod Delivery No Rx Phentermine
Phentermine For As Low As $6.00
Phentermine No Script 30mg
Purchase Phentermine Check
Phentermine Adhd Drowsy
Cheap 90ct 37.5mg Phentermine
Difference Btween Phentermine And Meridia
Buy Phentermine With American Express
Buy Phentermine Without Presc
Phentermine Hcl 37.5
Buy No Phentermine Prescri Ption
Snorting Phentermine
Cheapest Phentermine 90 Day Orders
Phentermine Direct Net Contact
Phentermine Overnight No Prescription Legal Legal
Phentermine And Alcholo
Phentermine Canadian Pharmacy
Phentermine In The Uk Only
Pravachol Actos Phentermine Actos Tysabri
Buy Phentermine In
Phentermine Under $100.00
No Primary Care Physician Phentermine Source
Phentermine 37.5 X 90 Tablets
Phentermine Heart Murmer
Phentermine Now
Cheap Phentermine 100 Tablets Overnite Shipping
No Prescription Required Phentermine Online
Drug Phentermine Buy Online Phentermine Uk
Phentermine Online Purchase From Namibia
Phentermine Erection
Phentermine 2037.5
Phentermine Overnight Paid By Mastercard
Phentermine 37.5 Mg Free Shipping
C D O Overnight Phentermine
Drug Phentermine Affect Y
Phentermine Shipped With 10 Business Days
Phentermine 37 5 90 Sale
Ordering Phentermine Phentermine Online Without Prescription
Phentermine Without Prescription No Doctor
Order Cheap Phentermine Online Without Prescription
Phentermine Web Pages
Information About Phentermine Hydrochloride
Phentermine Diet Pills Diet Pills
Phentermine Information Pharmacy You
Phentermine 37.5 Next Day Delivery
Phentermine For Add
Out Of Control Price On Phentermine
Buy No Perscription Phentermine 375mg
Phentermine With Satruday Delivery
Phentermine Prescription Caltran Diet Pills Phentermine
Order Phentermine No Prescription Needed
Doctors That Prescribe Phentermine Ontario California
Phentermine Without Dr Order
Phentermine Onlien
Purchase Phentermine Purchase Phentermine
Cheap Phentermine Diet
Buy Card Debit Online Phentermine Tramadol
No Prior Prescription Required Phentermine
30 Mg Phentermine No Doctor
Pharmacy Online Phentermine Online Description Price
Advanced Pharmacy Phentermine
Purchase Phentermine Phentermine Online Purchase Phentermine
Phentermine And Prostate
Phentermine 37.5 Online No Pescription
Discount Phentermine The Offical Site Index
Buy Phentermine 32
24 7 Phentermine
Phentermine 30 Mg Buy
Buy 30mg Phentermine Online Physician Consultation
Phentermine In Orland Park
Phentermine Next Day Ship
Natural Alternatives To Phentermine
Phentermine No Dr Approval
Purchase Online Phentermine No Prescription
Buy Phentermine Online No Prescription
Buy Phentermine No Prescription Needed
Buy Phentermine Cheap Without Prescription
Mexico Pharmacies Phentermine
Online Prescription Viagra Phentermine Meridia Adipex
Buy Phentermine With Next Day Shipping
Method Of Payment Accepted Cod Phentermine
Phentermine Eft
Phentermine Sibutramine Best Prices
Phentermine With Online Doctor Consultation
Phentermine Purchase On Line
Phentermine Pills Online
Where To Buy Phentermine 105 Mg
37.5mg Diet Phentermine Pill
Phentermine No Precription
Diet Pills Phentermine Prescription Adipex
Phentermine Tricks
Cheap Phentermine Without Prescription Free Shipping
Phentermine And Testicle Swelling
Cod Phentermine Overnight
Phentermine Shipped To Louisiana
Overnight Shipping Phentermine
Phentermine In The Uk
Low Carb Dieting Diet Phentermine Pill
Cheap No Rx Phentermine 30mg 90
Diet Pill Phentermine Huge Discounts
Phentermine Website
Cheapest Cheap Phentermine
Phentermine Mg Phentermine Discussion
Phentermine 37.5 Overnight
Atkins Diet Menu Diet Phentermine Pill
Phentermine Prices Mexico
Phentermine Online No Prescription Online Pharmacy
Phentermine Onlinephentermine Online
Phentermine With No Prescriptiom
Good Site To Buy Phentermine
Phentermine Metabolize
Actavis Mfg Phentermine 30 Mg
Phentermine By Fedex
Phentermine Hives Rash
Phentermine Green And White Capsules
What Does Phentermine Look Like
Original Phentermine Without Doctor Approval
Cheapest Phentermine No Presc
Drug Loss Phentermine Weight
Phentermine 37.5 90 Day Supply
Didrex Diet Pills Purephentermine
Can I Take Phentermine With Zoloft
Cheap Phentermine All States
Weight Loss Journal Diet Phentermine Pill
Cheap Phentermine 37.5 No Rx
Buy Scam Phentermine
Non-Prescription Phentermine
Order Phentermine Descriptive Phentermine Guidelines
Top Online Open Directory Phentermine
3pm Cheap Phentermine
Phentermine Yellow Pharmacy Online
Buy Phentermine Or Adipex
Phentermine 37.5 90 Free Shipping
Doctor Online Prescription Phentermine
Welcome To Cheap Phentermine Net
Phentermine Online Doctor
Phentermine No Rx
Phentermine And No Rx
Phentermine With Cod Charge
On Line Phentermine
15mg Phentermine
Phentermine No Prescri Hr Style
Phentermine And Sleepiness
Online Doctors Who Prescribe Phentermine
Phentermine Unsucessful Stories
Phentermine Online Cheap Phentermine Buy Phentermine
Celexa 2bphentermine
Overnight Us Phentermine
Phentermine 30 No Rx Fedex Cod
Weight Loss Or Phentermine In Kentucky
Innovative Pharmacy Phentermine
Phentermine Use Our Doctor
Order Phentermine Without A Prescription
Phentermine Vs Amphetamin
Oder Phentermine By Cod
Adipex Phentermine Guaranteed Lowest Prices
Phendimetrazine And Phentermine
Online Phentermine Physician
Authentic Phentermine
Phentermine Mg Capsule Phenterminechik
Does Phentermine Causew Skin Problems
Phentermine And Menopause And Dangers
Buy Phentermine With Mastercard Credit Card
Phentermine Grams Cheap No Doctor
Phentermine For Sale In The Usa
Phentermine With No Persciption
Phentermine 37.5 Sale No Prescription
Online Order Phentermine
Order Phentermine Online Without Prescription
Bulk Phentermine
Fast Virginia Phentermine
Phentermine Only Web Sites
Phentermine On-Line Without Rx
Phentermine Rr
Federal Express Phentermine
Phentermine In Jonesboro Arkansas
India Generic Phentermine
Phentermine Discussion Forums
Phentermine Low Prices
Phentermine On Line With Saturday Delivery
Phentermine Prima Pharm
Buy Epharmacist Phentermine
Phentermine Prescription On-Line
Cheap No Phentermine Prescription Prior
37 5 159 A Phentermine
Overnight Phentermine No Rx
Prescription Phentermine Without A Prescription
Phentermine Addiction Online Consultations And Prescriptions
Phentermine No Prescription Or Doctor Notification
Zoloft Phentermine Interactions
Phentermine 30 Blue 90
Buy Phentermine Online Pharmacy
Information About Phentermine
Phentermine Free Doctor
Phentermine Fda Approval Date
Phentermine No Prescription 30 Mg
Phentermine 35.5 Mexico
Phentermine Prozac Umaxppc
Best Online Pharmacy Phentermine Diet Pill
Best Phentermine
Online Phentermine Articles
Viagra Tramadol Phentermine
Phentermine No Preescription
Phentermine For Sale Phentermine
Lose Weight Phentermine Cod Orders
Weight Loss Phentermine 37.5mg Websites
Phentermine Weight Loss Drug
Difference Between Adipex And Phentermine
Phentermine Heart Problems
Phentermine Legal In Canada
International Phentermine No Rx
Online Pharmacy No Prescription Required Phentermine
37.5 Phentermine And No Prescription
Phentermine Without Doctors Consent
Phentermine 37.5 Free Online Consult
Weight Loss Clinic That Gives Phentermine
Discount Phentermine No Doctor
Phentermine 37.5 Without Prior Prescription
Phentermine No Presription
Phentermine Presciption Diet Pills
Cheapest Phentermine Cheap Drugs
Day Delivery Next Phentermine
Phentermine 15mg Overnight
Phentermine Results
Buy Phentermine No Prescription Up
Cheap Phentermine No Prescription 90×90
Phentermine 37.5mg 90 Tablets
Buy Phentermine Trap17 Net
What’s In Phentermine
Best Offer Diet Pill Phentermine Viagra
Counterfeit Phentermine Tablets
Cheap Phentermine No Doctor
Buy Phentermines Online
Phentermine No Presription Overnight
Where To Buy Phentermine In Australia
Phentermine And Pregnancy Test
Purchase Phentermine Purchase Phentermine Onli
Buy Phentermine United Kingdom
Phentermine Diet Pills No Prescription
Diet Plan With Phentermine
Mole Home Finance Phentermine Diet Pill
Cheap Phentermine Pills Pharmacy Online
Phentermine And Drug Testing
Phentermine Zoloft
Buy Phentermine Consult
On Line Prescription For Phentermine
Phentermine Consult Fedex
Rx Phentermine
Phentermine Odor
Buy Phentermine Information Phentermine Cheap Phentermine
Phentermine From Germany
Customer Reviews About Save Generic Phentermine

Oct 292008
 

A recent survey by McKinsey highlights the growing challenge faced by marketers in deciding how to allocate their media spending:

The rapid growth of online advertising hides a serious challenge: the digital world has developed faster than the tools needed to measure it…A June 2008 McKinsey digital-advertising survey of 340 senior marketing executives around the world shows the breadth of the gap between what’s needed and what’s available. Hobbled by nascent technologies, inconsistent metrics, and a reliance on outdated media models, marketers are failing to tap the digital world’s full power. Unless this problem is addressed, the inability to make accurate measurements of digital advertising’s effectiveness across channels and consumer touch points will continue to promote the misallocation of media budgets and to impede the industry’s growth.

There are three problem areas:

1. Media Planning – New tools are needed to help media planners compare the impact of on- and offline efforts.

2. Conversion measurement – Greater insight is required into how online messaging converts target consumers into making online and offline purchases.

3. Social optimisation – Targeting methodologies have not yet adapted to the changing context in which individuals are consuming online content; in particular, changed context(s) within social environments and how word-of-mouth and recommendations fit within this category.

The survey found that over 50% of respondents were not happy with the current processes for media allocation and measurement. Surprisingly, only 50% of respondents indicated that they used click-through rates to determine the effectiveness of their direct-response advertising – which suggests the rest are preferring qualitative over quantitative measures. Only 30% of respondents indicated that they considered the offline impact of online advertising.

Few in the online industry would claim that the ‘measurement challenge’ has been solved, but it doesn’t seem fair to suggest that poor metrics alone is holding the industry back. Despite having access to over a decade of data on how to use the online medium effectively (either as a stand-alone channel or in conjunction with other channels), it seems that a sizeable number of marketers have failed to adapt their toolkits and processes, or invested in the requisite skills to optimise the ROI of their online spend.

 

Oct 242008
 

Michael Heller has just released an ebook Gridlock Economy: The Tragedy of the Anticommons, which provides a short overview of the thinking encapsulated in his new book by the same name.

It provides an interesting counter-point to the growing trend of co-creation and  co-ownership of various assets, from (physical) property through to software, knowledge and IP.

Private ownership usually creates wealth. But too much ownership has the opposite effect—it creates gridlock. When too many people own
pieces of one thing, cooperation breaks down, wealth disappears … everybody loses. Gridlock is a free market paradox.

Heller argues that current wealth creation strategies – particularly knowledge-related (e.g. patents and copyright) – are highly dependent on the ‘assembly’ model: taking bits and pieces of previously created knowledge and assembling them in new ways. A case in point is modern drug research. New discoveries are almost entirely reliant on earlier research activities, some of which may be covered by patents. Similarly, when researchers find new applications for existing drugs, they will almost certainly need to secure permission from the drug’s owner.

Successful ‘assembly’ business models can be tough, but not impossible, when assets are concentrated in the hands of a few. However, when ownership is heavily fragmented and where, as is often the case, a single hold-out can scuttle a deal, then ‘assembly’ business models become highly problematic (hence the term ‘gridlock’).

The ‘tragedy of the anticommons’ arises when ownership of a specific resource is divided among too many entities, with the result that it is under-utilised or lay unused. One example from Heller:

Consider the example of a brother and sister who jointly inherit the family home. “All of us as parents want to believe our children will be friendly when we’re gone,” says an estate planning expert, but leaving the house to the kids is “a sure recipe for disaster.” One wants to rent the house out; the other, tear it down. If they can’t strike a deal, neither can move forward. The house sits empty. That’s gridlock…Now imagine twenty or two hundred owners. If any one blocks the others, the resource is wasted. That’s gridlock writ large—a hidden tragedy of the anticommons. I say “hidden” because underuse is often hard to spot. For example, who can tell when dozens of patent owners are blocking a promising line of drug research? Innovators don’t advertise the projects they abandon. Lifesaving cures may be lost, invisibly, in a tragedy of the anticommons.

Jul 302008
 

I recently had cause to investigate how you might use a utility/cloud computing service, like Amazon S3, to provide an efficient and modestly priced solution to serving up video-on-demand content services, as an alternative to the more traditional offerings of streaming video platform providers.

I was impressed by the level of intelligence behind not only the pricing of such services (just cheap enough so it makes more sense to outsource – Coase theorem at work), but also the dynamism reflected in the engineering of their underlying infrastructure.

Amazon, and others, provide some seriously ‘grunty’ utility/cloud infrastructure. The recent S3 outage notwithstanding, utility/cloud computing is here to stay. Its adoption by businesses is likely to be accelerated by the early successes of server virtualisation technology adoption by enterprise users, spearheaded by VMWare.

This points to an exciting new realm of opportunities for "middle men" to move this capability into the hands of consumers. Allow me to explain.

We are the last generation who will ever have to worry about launching an installation CD.

We are the last generation who will ever have to worry about disk (storage) capacity.

We are the last generation who will ever have to worry about CPU speeds.

Finally, but importantly, we are the last generation who will ever have to worry about bandwidth throughput (speed).

Imagine what you could do in a world when you’re free of these kinds of constraints.

We’re starting to see the potential indirectly, in the guise of the rapid proliferation of Facebook apps. These apps exist "out there". We don’t care where they reside. Importantly, we play no active role in their installation, configuration or management. Equally, these apps care little about the end user equipment/device or infrastructure being used to access and interact with them.

This is how software (and services-powered-by-software) will look for all users within the next decade.

The promise of utility/cloud computing infrastructure, then, is that it will finally hide the technology. And history tells us that when technology ceases being ‘technical’ – when technology shifts from the core of the experience to the periphery – immense behavioural changes follow.

Now back to the opportunity for middle men.

While utility/cloud computing infrastructure has hidden (or, at least, is starting to hide) the technology, it is still a fairly immature offering, in that – as a product offering – it is largely focused at technical users.

While it is easy to setup an Amazon S3 account, for instance, it is still quite difficult to configure and manage the services that you purchase/consume. This represents an unnecessary hurdle in the current environment, in which – thanks for ‘mashup’ services and the like – non-technical users are being increasingly encouraged to ‘develop’ software-based solutions to every day problems.

This creates a market for an intermediate layer: companies who take utility/cloud computing services and repackage them in a way that makes them accessible to non-technical users. Companies like Morph have taken a step in the right direction, but there is still plenty of scope for innovation in this new market niche.

Jul 182008
 

I just had a quick scan through a  report by JP Morgan’s Internet analyst Imran Khan, titled Nothing But Net, released in January 2008.

In an almost throw away line, he notes that "[m]ore than 80% of online inventory currently sells for less than a $1 CPM," with an average CPM rate of just $US3.31 for display advertising in 2007.

As indicated in the two charts below (the first showing the global market, the second the US market), while rates are forecast to lift in coming years, they won’t do so by any order of magnitude.

 

image

image

I found these figures quite surprising, given that in several recent projects, clients were forecasting (and, based on trading history, expecting to secure) CPMs ranging from $A30-$A70.

There are a number of contributing factors that create pressure on CPMs for display advertising, with the primary two causes being:

- A glut of non-premium advertising inventory (as mentioned previously, sites like Facebook are generating so much traffic that it completely devalues their inventory)

- Low adoption of ad targeting platforms and other revenue optimisation techniques.

But how did it come to pass that display advertising rates have slumped so low?

The rise of ad networks, which aggregate unsold/distressed advertising inventory and sell bulk quantities at low rates to advertisers who are happy to run "spray and pray" campaigns, is certainly a factor.

Another factor is that marketers expect a much higher correlation between digital marketing activities and confirmed sales than they do in any of the ‘offline’ media, which I have also covered previously. This is quite ironic, yet by viewing online marketing as undifferentiated from other types of direct response campaigns, marketers have driven CPM rates down.

However, the main reason why CPM rates for display advertising remain at such artificial lows is that publishers (and the online industry in general) have still not developed a compelling story for advertisers when it comes to achieving ‘best practice’ for display advertising in online brand/branding campaigns.

With the dominant mindset among marketers being that the online environment is a sales-focused medium and, consequently, the aim of online campaigns is to secure reach, they buy based on tonnage: how many impressions will you deliver for my budget?

Publishers need to start the process of educating marketers about engagement and, in particular, the types and degree of engagement that can be achieved within an online, interactive environment that could never be achieved in other, display-like mediums (such as newspapers, magazines and TV).

Importantly, publishers need to get marketers thinking about how online advertising can be better utilised to achieve brand awareness, interest and preference. Marketers need to recognise why online advertising should be part of the marketing mix at all stages of the purchase cycle, rather than only at the ‘buying’ stage.

The first step in this process is to get everyone on the same page as to what ‘engagement’ is and, importantly, how you can tell whether you have it (yes, another metric!).

The second step is to recognise that the ad units themselves need an overhaul. Display ads have been around since the very first Web browsers were released (some 15 or so years ago). This might lead you to think that they are now a ‘mature’ product. But display advertising is still a very young format, when you consider how long print and TV advertising formats have been around. There is still plenty of opportunities to refine and evolve the advertising formats themselves.

Lastly, publishers need to recognise the importance of creating and sustaining their own brands. Having a well-defined, understood and trusted/respected brand is key to achieving engagement with audiences. Furthermore, publisher’s brands create the opportunity and the context for introducing advertisers (and their brands) to the audience.

All too many publishers fail to adequately engage with their own audiences and nurture their relationships. Until they solve this part of the equation, they will be hard pressed to improve advertisers’ perception of value for their advertising products.

Jun 282008
 

In the past three years, we have witnessed emergent signals of a radical change in the nature of the Internet. Like the transition that preceded it, where its focus shifted from its original, technological orientation to a content-centric positioning, this next phase will bring a dramatic shift in fortunes for some of the major players.

While subtle, the change is profound. The label many use to describe it – the Come to me Web – best captures its essence. To appreciate the ramification of the changes in both technology and consumer behaviour we are witnessing, we need to revisit its precursor: the I Go Get It Web. Under this paradigm, content was at the centre of the universe, and most effort was directed to targeting and luring consumers to the content. The balance of power lay in the hands of the content owners; the leading business meme advocated aggressively capturing and monetising ‘eyeballs’.

In this environment, consumers were obliged to locate and seek out content and related services of interest. Not surprisingly, search engine technologies were highly valued, as were portal-style offerings that optimised the effort-reward equation by aggregating useful content in a single place.

The dominant mindset of content owners was to publish content for a single purpose – reading – and to treat computer screens as the digital evolution of printed books. The business objective, then, was creating and disseminating read-only information, and the challenge was making that information findable.

In the I Go Get It world, ‘findability’ became the overarching metaphor, and it borrowed much from the physical business of publishing, as evidenced by the emphasis placed on replicating print layouts and making content navigable through the use of publishing concepts. That most consumers accessed content via a single device undoubtedly influenced this state of affairs. It led to ‘format’ wars, as content owners sought to lock-in consumers by mandating the use of plug-ins and other proprietary formats.

The adaptation of a physical product mindset brought with it a range of artificial constraints around content access, use, and re-distribution and, similarly, how to enhance content products via the connectedness the Internet enabled. Only now, as our understanding of the true potential of digital services matures, and as emerging technologies decentralise power, are we seeing a dramatic change in both our metaphors and use of the Web.

The Come to Me Web neatly inverts the paradigm. Users, not content or technology, are at the centre of things. They are using tools and services that bring the content to them, ready to hand, on their preferred access device.

Decoupling content from its publishing containers has enabled consumers to use, aggregate, store, re-use, re-combine, and re-distribute content. This has fundamentally changed consumers’ perception of content: it has become “my” news headlines, “my” music, and “my” weather. This changing usage, and altered expectations, has forced the growing adoption of open standards and formats, as consumers seek to ensure maximum availability of ‘their’ content. Service owners are even starting to promote open APIs, so that others can create useful tools using their content, because this is what consumers are demanding.

The main hurdle for consumers has shifted from being able to find information to being able to re-find information they have already collected. This has spawned increasing activity in the area of ‘ambient discovery’; technologies designed to sniff out clues about what you will be interested in next (as distinct from now) and present you with it before you even ask.

These changes in consumer behaviour will have significant impact on the business of the Web. We can expect to see two things: increasing audience fragmentation and decreasing site traffic. Both will remain obscured initially due to an overall rise in online activity driven (ironically) by the increasing usefulness of new information platforms.

Increasing audience fragmentation will occur because consumers will become more responsive to site and content recommendations provided by peers and new technologies, displacing the primary role of search engines in this area (which tend to concentrate traffic on a handful of popular sites). Site traffic will decrease because consumers will be able to avoid visiting the original source of content, either because the content is coming directly to them (say, via RSS) or a peer is passing it on. Given the predominance of advertising-driven revenue models, Web measurement metrics will soon evolve to address the obvious difficulties this creates.

In the Come to Me Web environment, it will be incumbent upon service providers, and advertisers to go where your audience is, not where you would have them be.