LATEST ENTRY

MARKETING | Noah Brier

The Math of Viral

Trying to use power laws and Metcalfe's Law to understand how viral marketing actually works.

June 1, 2007 | RSS | EMAIL | PRINT | 10 COMMENTS

I've been working on this thing in my head for a while and it's still hardly played out. But I've got a thesis going that you can apply Metcalfe's Law to marketing. For those not familiar with it, Metcalfe's Law is a way to calculate the value of a network. In a recent Forbe's article Metcalfe himself explained it like this: "The network effect says that the value of that Ethernet card to the person on whose desk it sits is proportional to the number, N, of other computer users he can connect to. Now multiply this value by the number of users, and you have a value for the whole operation that is roughly proportional to N2." In plain English: If one person has a fax machine it's got no value, if two people have it, it's still got very little value. From there, however, the value really starts to grow for everyone involved and eventually it hits a 'tipping point' where you can't not have one.

metcalfeslaw.gif

[Taken from Metcalfe in Forbes]

As I've mentioned, power laws are a core part of our interconnected world and Metcalfe's Law speaks to how these things happen. Essentially, because there's more value to a network as it grows, eventually they reach a 'tipping point' (or 'critical mass crossover') where some serious growth kicks in (think about when everyone started texting you and you had no choice but to text back).

[WARNING: I am probably about to butcher some math. I apologize for that.]

Anyway, when I was looking at that graph of Metcalfe's Law, I realized that it's exactly what 'viral marketing' is attempting (as much as I hate that term). Essentially 'viral' (and PR) is an attempt to get value that goes above and beyond cost. In the old days of media buying you bought the number of impressions you wanted. It was a fairly straightforward equation: Value = cost, and cost is based on number of impressions (N) times CPM (cost per thousand impressions). 'Viral' tries to beat the system, however: By tapping into the network effect a marketer can gain far more value than cost. That's because while cost is still a constant, value grows exponentially based on the number of people who the 'viral' touches (that is assuming it is spread-worthy).

[Deep breath.] Okay, that was way more math than I have thought about in ages and now my head hurts. But I think I had to go through it. I don't think anything I said in the last paragraph is particularly revolutionary, it's more just putting some structure around things lots of people have talked about/noticed. What is more interesting to me about this whole thing is the idea that if you lower the cost, you actually lower the 'critical mass crossover'. Basically, 'going viral' is not an absolute, rather it's based on the cost of production. If you make something for $5 than the 'critical mass crossover' is far lower than spending $500,000. This is possible because all power laws are scale-free, meaning no matter where you look at them they're exponential growth.

The way I've been explaining scale-free to people lately is with exponents. 22 = 4 and 42 = 16 and 162 = 256. When you look at those four numbers (2, 4, 16, 256) you think that exponential growth kicks in when you square 16, I mean the jump between 16 and 256 is huge. However, if you just isolate 2, 4 and 16, the jump between 4 and 16 is also huge. Below I've put a few charts, while each jump is more radical than the one before it, they are all radical in their own right.

exponential.png

Okay, one more point to make and then I'll leave you alone. If rather than approaching 'viral marketing' with one message, companies tried a number of different, low-cost options, then they might find more success. Each would have a lower 'critical mass crossover' point and you'd only need one success to pay for the rest.

Of course all this assumes a marketer has made something people actually want to pass around, which in and of itself is a lot to ask . . . (For more thinking on this topic I suggest this paper on Big Seed Marketing by Duncan Watts [PDF])


PREVIOUS ENTRY | NEXT ENTRY

LEAVE A COMMENT

First name, first and last, whatever you feel like.

Required, but not displayed (so don't worry about spam).

If you've got one, flaunt it.

You can use some HTML (a's, br's, p's, oh my!) if you'd like, if you don't know what that means, don't worry about it.
REMEMBER ME?

COMMENTS

1Josh

I don't think you butchered any math at all, man. That was great.

Have you heard that metaphor about how if humanity's existence was represented by a clock face, the last 30 seconds or so would show vastly more technological developments than the rest of the time in its entirety? I've wondered for a long time if humanity's growth and development didn't take place basically at a linear, arithmetic rate until electricity came along and it tipped over into an exponential/geometric rate. I mean, I guess it must have. That's no amazing relevation either, but it's the kind of thing we need to think about, because our culture (like any culture) invokes the past all the time for various reasons, and our present is vastly more different from our past than was the case for any previous generation.

Material like this post makes me really hopeful, Noah. I have a lot of faith in math (even if I stopped taking it almost half my life ago), and I think there are rules we can determine about our group behavior and that we're better off figuring them out.

June 1, 2007

2Peter Caputa

I like the math. The engineer in me is happy to see this.

And I agree that marketers should spend their time trying different approaches.

Taking that a step further, shouldn't they be building their own network of contacts through genuine interaction and interest in other people? Rather than trying to create something funny/racy/etc that someone else might tell a friend?

June 1, 2007

3Josh

Taking that a step further, shouldn't they be building their own network of contacts through genuine interaction and interest in other people? Rather than trying to create something funny/racy/etc that someone else might tell a friend?

Peter, I think that when you create a piece of applied art (like a commercial) that works, it's for the same reasons that a piece of arty art (like a book or a movie) works: because you're saying something true—true in the sense that it resonates with people, anyway. The difference (and the problem), course, is that a novel just has to be true in the resonant sense to work; an ad that works is actually creating faith in a business, and the business may not deserve it. It may just have been smart enough to hire a good agency.

Anyway, contact with people and genuine interaction and interest are good things, but if somebody can find away to make me aware of a worthwhile product that its maker stands behind, frankly, I don't care if they make Emily Dickinson look like a social butterfly. ;-)

June 1, 2007

4Max Kalehoff

All along, I was thinking: "Noah, you've got to read Duncan Watts' paper," and then at the end, you did. In a marketing context, I think Metfalfe's law falls way short because it just doesn't account for qualitative factors; it seems to assume all things equal. I steer more to Duncan Watts' theory, which uses biological viral spread as a metaphor. Duncan Watts' underscores the naivety and often stupidity of so many marketers and agencies by placing their eggs in "standalone viral marketing strategies" baskets, because the fact is that nearly all fail. Rather, he argues, larger marketing and media plans should incorporate viral strategies to achieve multiples of success. For example, by viral standards, if you contacted 10 people and only five people passed along a marketing message, that would probably be a dismal failure. Conversely, if you subscribed to a traditional push-marketing impressions plan, where five of those ten people passed along a marketing message, then that would equate to a 50% addition lift on the face-value impression buy. Of course, that's a grossly simplified scenario, but makes a lot of sense. For one, it begins to provide some justification for Super Bowl advertising, at $2 million+ per 30 seconds of airtime. I don't mean to go engagement on you, but if engagement is the context that results in optimized results or achieved business objectives, then viral strategy should be embraced. I've been preaching this concept for years, but it seems to be coming up through this context here.
Cheers,
Max

June 1, 2007

5Noah Brier

Wow, awesome comments. Let me go one at a time.

Josh, thanks so much. I think that metaphor is perfect and certainly speaks to power laws. Have you read John Hagel's Power of Power Laws? It's one of the best things I've read in recent memory and speaks to (indirectly) why this speed up may have happened. Essentially power laws show up when things are connected and the last 500 years has been about technology (including electricity, the internet and things like cargo ships) that connect the world.

Pete, I think you're right, marketers do need to find ways to build a network, just like people do. As I understand it, this is one of the things Tremor does. I'm sure Max would have some interesting thoughts on this one as well . . .

Max, you are absolutely right and that Watts paper was an excellent explanation of something I think most of us recognized before. I was mostly trying to put some additional framework around 'viral' and explain some different approaches (all assuming you have something interesting to say). But I totally agree with you, we can't ignore the interconnectedness of all this media (and probably never should have). When I was writing this, actually, I realized the same math probably applies to something like PR . . .

As a side note, another question I have is how do you attach value to bottom line? Anyone got a good answer for that one?

June 2, 2007

6amber

whoa...this was the perfect thing for me to read today, as I sit here and write briefs for...viral campaigns. :)

The value of "viral" campaigns is such a hard thing to measure - even if the campaign is a success in that exponentially more eyeballs than you first put it in front of see it, how do you measure it's success in actually selling something?

I know brand awareness is certainly valuable in itself, but what do you do when you have a viral campaign that can be considered a success because it spread to a large quantity of people, but in the end didn't yield any measurable results for the product?

Ok, I'll be more specific - I worked on a campaign for a movie, which included a "viral" video component along with a traditional print, TV, OOH, and online ad buy. The viral pieces weren't just thrown online while we prayed they would take flight; instead they supported a large, integrated campaign - this is what Watts is talking about with Big Seed Marketing, right?

In terms of viral spread, the videos did well - more than half a million views on YouTube within a week. The ad campaign performed ok too - no dismal numbers, really.

But guess what? no one went to see the movie!!! the box office numbers were horrendous - there were more views of one viral video in the campaign than tickets sold to the film.

I guess that means that the value we did get out of the viral spread didn't really translate when it came time for consumers to act upon the messaging they'd seen ( and apparently enjoyed enough to pass it on). I suppose if they all buy the movie on DVD when it comes out, then that would be different, but I somehow doubt that will happen.

Sorry - this was a really long comment! But what do you think? Is there a way to know what percent of all the people talking online, and passing videos to friends, and engaging with the brand digitally actually spend their money on it's products? Again, that's a lot of brand awareness you're getting, but if "influencers" aren't buying, and the people they tell aren't buying, how much does awareness really count?

June 2, 2007

7Herb

Amber - First thoughts...was the movie actually any good? A box with crap in it is still a box with crap in it...no matter how shiny the advertising. I'm sorry...that was not meant to be mean.

I wonder when we will start seeing 'engaging viral' that is more than just asking someone to pass something on. Than you could have a measurement of engagement. For example, at the very basic level, not only did I pass on the 1 minute video but I claimed the coupon and bought the 50% off soda (not the funnest type of engagement, but engagement non the less).

Mass viral worked because hey who doesn't think seeing someone take a ball to the nads is funny? But it doesn't really resonate other than the shock/funny value. Yes that f'in chicken was viral and got a ton of views and may have gotten people into the brand, even got the brand into a conversation or two because of it...but how come we never hear about how many sandwiches it sold? (if it did and I missed those blog post please let me know...and if you ask me, the only people who actually watch that thing were people in the industry). What I think we need to start seeing more of is contextual relevant viral or target relevant viral, not mas viral.

With contextual relevant or target relevant viral you could add an engaging component that would encourage interaction with likeminded brand consumers, brand itself, or the product besides just passing the viral on. I've got more thoughts on this, but I'm so tired I can barely type (2 soccer games today is killing me). But having something engaging is how you get your influencers involved, probably easier said than done.

Noah - ever wonder why a comment doesn't have their own comments? So you could branch out on a particular comment and continue side conversations? Or is that just to meta even for bloggers?

June 4, 2007

8amber

Hey Herb,

to answer your question ( and i'm not offended at all, and completely agree) - the movie was actually deemed a critical success, got good early reviews in the trades, and the actors in it got favorable reviews - it just bombed at the box office. go figure.

it's funny - i'm having this same conversation in real life with my coworkers, and you're all making a lot of great points that are totally helping me out. so thanks!

June 5, 2007

9Ian

I might be [ok, I for sure am] over simplifying Watts' argument ... but I see his 'big seed marketing' concept as glorified spam.

June 5, 2007

10Craig Stanford

V = 1:1.x

May 15, 2008