The Problems with Delivering Real Goods from a Virtual World


I spent another great session at the VIO seminar in Second Life on Monday morning (my time). Xander Newman covered how to deal with real life businesses that are using Second Life for vCommerce. All great stuff and it, deservedly, drew an awesome crowd. However, at the usual after-seminar discussion someone mentioned that they wished they could buy pizza using Second Life. Buying real life goods from a virtual world is not a new thing. Both Dell and American Apparel tried it in Second Life back in 2006. By all accounts both were dismal failures. Likewise delivering pizza (or any other real life product) faces some monumental hurdles to make the experience both worthwhile for the retailer and a valuable alternative to the comsumer.

Believing that something mundane is cooler, better, more elvish when done in Second Life or any other virtual world is hardly a new phenomenon. Back in the late eighties it was considered cool to use the command line to trigger the Coke machine to deliver a can of Coke at the local university computer club. In the nineties it was considered cool to order pizza over the web. The geek-set in particular are known for pushing the boundaries of technology in order to achieve an otherwise mundane result. There’s nothing wrong with that, in fact pizza ordering over the web has become a major marketing differentiation point for some companies. One that their competitors have rapidly employed as well. However doing something for fun and making it into a viable business solution are two very different things.

Consider the problems associated with attempting to deliver pizza from Second Life:

  1. Pizza has a limited shelf-life period and so must be provided locally (usually within 30mins or it’s free =). Second Life is global. So any pizza franchise offering orders from Second Life must have a worldwide network of stores. The only pizza chain I know that has anything resembling that is Pizza Hut and they don’t even have a global ordering system for their web interface
  2. Most pizza places are open (and deliver) only during limited hours. Second Life is open 24/7.
  3. There’s nothing linking your Second Life avatar to you and your delivery address. Therefore the potential for fraud is high. Especially with non-pay info accounts.
  4. The Second Life populace is extremely small in comparison to other (and more useful) ordering options like the web or telephone so the ROI is likely to be comparatively small.

One option that may work as a marketing solution is linking a Second Life object to a web interface so that clicking on the object brings up an external web site where you can order your pizza. That doesn’t get around the first two hurdles mentioned above but it may be worth examining as option for a major pizza chain. Certainly the initial outlay would be minimal. However, it may still be non-viable until Second Life has better browser implementation.

So, if Pizza is problematic then what about other real life goods? One option that at least initially looks appealing is the furniture market. An organisation like Ikea who have a worldwide distribution network could conceivably benefit from having a virtual showroom where consumers could see their virtual items on display, buy the virtual item and then try them in a mock-up of their real life home or office.

However even here we’re faced with hurdles that may make the idea non-viable:

  1. Second Life’s modelling features are very limited. Getting exact models from an external source and importing them into Second Life is difficult. Second Life only accepts a small number of modelling formats.
  2. Even when you’re able to import the models, Second Life may not be able to render them exactly as they appear in the real world. This may create potential legal problems depending on where you are delivering the goods. A disclaimer mentioning the possible differences between the virtual and real-world objects may suffice but what if it doesn’t? It’s something that needs considering during the planning stages
  3. At least in Ikea’s case, different stores sell different goods and sometimes at different prices. This means that ordering in Second Life and having the local store deliver it can pose a problem.

Delivering real world goods using a virtual world interface isn’t impossible, but it should never be done just because you can. The questions that always needs to be asked are why is this better? Why are we doing this? How is this helping us? What makes this better for the customer? Only after you have answered them should you consider implementing it.

About the Author: skribe

Based in Perth, Australia, Antonio Barimen (aka skribe) is a writer, digital media consultant and social media producer.

He is available to help you develop social marketing and digital media strategies, improving communication between staff, partners and suppliers or just increasing the number of fans on Facebook. He has developed successful digital and social media projects for clients including CBS, Evian, Procter & Gamble, Discovery Networks, Pernod Ricard and American Express.

Connect with him on Twitter or Google Plus.
Tags: , , , , , , , , ,
Post Under Advice, Business, virtual worlds January 28, 2009

19 Responses to “The Problems with Delivering Real Goods from a Virtual World”

  1. Couldn’t agree more. The only advantage for some of the companies who’ve done this is the PR value; practically, it’s a nonsense.

    You’ve commented in the past on Linden Lab’s obsession with making Second Life a business tool. It has applications in video, team-building and limited conferencing, but with very few exceptions, it is not a sales channel.

    • skribe says:

      I’m more than a little surprised that there are still some large companies that are still considering this. I think the evidence is pretty conclusive that you need to think carefully about selling real life goods in a virtual world. Know of any success stories?

      Also, apart from the issues I listed above one thing I didn’t include is the psychological factor that says that everything in SL must cost less than the real world. I encounter that on a regular basis when I get told that I should be charging hundreds of Linden dollars for a video that is going to take me 10-15 hours to complete. Any RL business that is selling anything real in SL needs to be prepared that their RL goods are going to be priced very high in comparison to the micropayments that the residents are normally used to.

      Thanks for the comment, Bret.

  2. The ‘Pizza Hut’ demo appears with each new generation of technology, and never fails to get laughs. We saw a web version. We saw a data-enabled/GPS-enabled mobile phone version. Now we see an SL version.

    It’s vaguely silly, but not, I think, for the reasons you say:

    – Second Life’s “globality” doesn’t disquality it here – any more than the phone system’s or the internet’s ubiquity disqualifies them. It would only disqualify a medium where “reaching the whole world” cost more than “reaching a neighborhood,” and in none of these media is that the case.
    – Second Life’s round-the-clock availability, similarly, isn’t a drawback. The medium is available when the store is open.
    – Linking an SL avatar to a real identity is easy. LL could act as a payment proxy – the same paradigm as when you use a mobile phone to buy stuff and the charges appear on your phone bill. Alternatively, a trivial registration process with a web-page pop could link your avatar’s UUID (a unique identifier) to a registered payment method.
    – Sure, the population is small. Which is really why this is a vaguely silly demo. But in fact, the cost of building a Pizza-Hut ordering system for Second Life would be negligible, particularly given that they already have a web/commerce interface. This is oft-trod technical landscape for competent SL/web developers.

    Likewise, quibbles about model accuracy, diverse store policies, etc., seem (to me at least) to miss the point. Ten years ago, critics of ecommerce said that people won’t ever order clothing on the web because web photos are small, grainy, and misrepresent color, pattern, size and material quality. Now, people order stuff on the web based on crappy photos all the time. All it took was a generational shift. And the trend (ecommerce is now booming) is now being accelerated by counterintuitive applications such as social media product recommendations — it doesn’t really matter what the photo of the stereo looks like if 51,247 customers voted it a Best Buy. Likewise, diverse store policies, tax regimes and other ‘cross border’ complications can be aligned by policy and compliance engines that can now be purchased on the open market. This is not the big, scary complex thing that even some developers still think it is.

    I suspect the real reason nobody’s done it is because … nobody’s done it. For those contemplating doing it, here are some further suggestions:

    – SL is not the audience delivery mechanism: The question isn’t whether your customer is in SL in significant numbers. It’s whether you can draw them into SL from a website whose traffic is, in turn drawn in by classic SEO, Google Adwords, email and other methods of connecting with cheap, quasi-qualified eyeballs. Any commerce project that doesn’t integrate SL with the web and take responsibility (on the web side) for audience generation, registration, qualification, and asynchronous community maintenance is a lose.

    – Since you can’t go broad, go deep. Instead of trying to sell spatulas and cheese graters in SL, think about selling the products that cost $25 million plus, and have six-month to three-year sales cycles requiring dozens of people to fly from place to place again and again and again to sit in meetings and educate, deliberate, collaborate to advance the deal. Sell green data-centers. Sell massive public works projects. Sell airliners. Sell custom yachts. Sell farm equipment in bulk. Sell chemical process plants. Or get a firewalled sim (easier than you think) and sell military stuff.

    The sales that really drive the world economy all live in the zone that SL excels at. Closing these sales means: 1) forming relationships, 2) building trust, 3) establishing expertise, 4) teaching (often through visual simulation), 5) learning, 6) aligning far-flung stakeholders, and 7) collaborating to “get to yes.” All this can be done really, really easily and cheaply in virtual worlds.

    – That said, there are SOME classes of low-cost consumer good (and consumable media) that carry with them a natural social resonance. Music sells on word of mouth and friend networks. Clothing (for particular demographics, e.g., teens) does too. Certain kinds of consumer electronics. Books. And by extension, probably kitchenware, tools and do-it-yourself stuff, etc. All this stuff sells better when embedded in a community and a conversation. The real world – and brick-and-mortar stores and malls – form one pole for these ‘social shopping’ networks. The flat web and socially-enabled e-commerce is another. And virtual worlds could easily be a third. You figure we couldn’t fill a sim (or four, or maybe eight) each Saturday if we did a Home Depot-sponsored do-it-yourself show in SL, maybe with a sideline focus on the dream-homes that people build for themselves there? And that would let Home Depot hit a nationwide/global audience for the same or less money than they could produce a similar event (“Come meet Bob Vila! from This Old House”) for several hundred people at one local store.

    John Jainschigg’s last blog Patents: First, They Came for your Word-Wrap

    • skribe says:

      Thanks for your comment, John.

    • Interesting comment about large-ticket items but the problem is this: business is time-poor. Purchasing Managers have no desire to spend a dozen hours in a virtual world learning the ropes in order to evaluate one competitor’s aeroplane.

      When you are selling something you want to make it easy for the purchaser, not difficult. Not contingent on their graphics card. Not contingent on their firewall. Not contingent on them spending lots of hours learning the interface. The user experience is not sufficiently compelling to compensate for these hurdles; buyers will not bother.

      It’s up to the seller to make the case in a format that is widely and easily accessible to the buyer. That’s video, text, personal presentations and the web.

  3. on delivery of goods – metaverse bundle time to comment

  4. yet another reason for mesh import into SL. But if I ever order a pizza from within SL shoot me.

  5. RT: @epredator: on delivery of goods – metaverse bundle time to comment

  6. epredator says:

    The only way we get to the best solution is through some less than idela ones. Without the attempts or demands to use certain technologies (as John points out with ealry ecommerce) we are not going find the patterns, models and business ideas that work.
    Virtual worlds are not just about another product advertizing medium in the traditional sense. However they can be a component. They can also incorporate other people, that is actually way more important than the particulars of the visuals.
    It should be easy, accessible, quick when needed to be quick, in depth when needed to be indepth. Various solutions will solve each of those attributes, but just as there is not one website answer to engaging with people and products there is not one virtual world one.
    Also you should not forget the important potential of using the virtual world as the design and distribution channel. Software is virtual already, as is music and phone ring tones. There is value in 3d content and talent in creating it. We then also have the exciting extra of being able to make the virtual real again with 3d printing/rapid fabrication. Again, its not perfect yet, but you can produce what you need when you need it, not mass produce things in remote places and ship inventory around the world. The virtual world platforms can act as a catalysts for this dramatic change in manufacturing.

  7. jon himoff says:

    more interesting perhaps is what things would work well from a virtual online environment. I think the virtual experience and micropayments can accelerate ad hoc, impulse buying. Issue is that you need to know the real human logistics info. We are now selling a real world t-shirt for virtual world people, but that is a web thing not inworld because we need shipping details. In fact, we hear that a lot of people do not want to link their avatar to their real person details on a 3rd party system for privacy…er…lack of privacy concerns.

    jon himoff’s last blog post..Why Do We Call It "Black Swan"

    • skribe says:

      That’s what we’re hearing too, Jon. Privacy is a huge issue. One I think will pose a major dilemma for anyone dealing with trans-national data transmission (which is just about every web site and virtual world out there) over the next few years.

      Thanks for dropping by, Jon.

  8. Peter Quirk says:

    The IKEA idea has a lot of merit, but I foresee the corporate lawyers inserting a click-through EULA on every piece of furniture to warn me not to copy the IP in any way.

    Peter Quirk’s last blog post..Sensor networks and virtual worlds

    • skribe says:

      It does, Peter, but it faces major obstacles and Second Life may not be the best place to do it. It might be better for Ikea to adapt their own browser-based world. Given the huge inventories involved that would require possibly thousands of hours to convert to something that can be used in SL you’d want as many people to use it as possible. Second Life doesn’t have the population nor the low-barrier to entry to warrant such a major investment imho.

      An EULA may suffice, however governments tend to be more strict about hardware than they are with software. Consumer Protection won’t tolerate a ‘buggy’ baby’s chair that occasionally -usually without warning – crumbles to a heap and requires reassembling, but the equivalent is accepted practice in the software world. Outside the US bait-and-switch is usually penalised very quickly.

      Thank you for your comment.

  9. […] answer our question, Srkibe, the blogger at Skribe Productions out of Australia, looked into the V-Commerce this very question, looking at pizza delivery originating in a virtual world. The […]

  10. Kiwidude says:

    I always thought the American Apparel idea had merit. Clothing seems like an ideal product line due to the way in which a Second Life identifies with their avatar.

    I myself bought a suit in RL that matched an SL outfit once, because after wearing it in SL I knew I would be comfortable with it in RL.

    However you’re right Skribe, there are still many issues to overcome such as world-wide delivery, local pricing etc.

    Ultimately I think we circle back to the web, and the best a retailer could do would be to provide a seamless link to an online order form (with a page of disclaimers and restrictions no doubt).

    Thanks for raising this topic, it’s aways good to ponder.

    • skribe says:

      Clothing has the same problems that any building does. You have to make compromises to be able to import them into SL. And seriously, how many SL avatars match their real life bodies? What may look good on the 2m size 6 avatar probably isn’t going to work as well on the 1.6m size 16 player.

      I think we use what works the best, if it’s the web, use that. If it’s some other medium, use that. That doesn’t mean we shouldn’t continually experiment and reinvent uses for alternate mediums but we should do so with a definite plan and a well considered set of options that deal with the hurdles – not just because its cool and because we can.