The Mathematics of Word of Mouth Campaigns

Alan Mitchell writes in Marketing that "Word of Mouth is over-hyped".

 

Voltaire once responded to the statement that "life is hard" by asking "compared to what?"

 

So can we mathematically compare how over-hyped WoM campaigns are to say advertising,  a historically over-hyped but also over-maligned benchmark?

 

Alan pivots his argument on the hard data of a Duncan Watts study on Twitter chains that found that "98% of tweets didn't cascade at all". I would want to dig deeper on what constitutes a "cascade" because the promise ("hype") of WoM is that messages will go viral, pass a tipping point, cross over into the main stream etc. not just be "passed on a bit".

 

But this data point does give us something to work with as it suggests a WoM success rate of 2%. 

 

Now at face value this compares very poorly with advertising that might (big assumption here) have a success rate of 50% on the basis that the much quoted Lord Leverhulme suggestion that "half my advertising budget is wasted, I just don't know which half" seems to have become a rule of thumb of the marketing community.

 

But the WoM community would respond (and I think i got this metaphor from Mark Earls, the Herdmeister) that "you do have to light lots of forest fires to create a conflagration but only one needs to take hold for you to succeed in your goal".  So best practice in WoM campaigning seems to include putting lots of "virals" out there and assume that at least one will catch on (fire).

 

Having said that, if the failure rate of your WoM initiatives is the Watt's average of 98%, some maths shows that you would need to put out 34 messages to give yourself an even chance of at least one spark being successful (and hence match the advertising benchmark).

 

But as a WoM - drum roll - "Practitioner", you might well claim that your messages are some what better designed to go viral than your average tweet. 

 

If you claim that your virals are twice as infectious (only 96 % failure rate) as the average then you still need 17 of them to hit the 50% success rate advertising benchmark.

 

If you claim to make virals that wash five times whiter (90% failure rate) then you need 7 of them.

 

If you think you are ten times tougher (80% failure rate) then just 3.

 

(BTW If you think that your virals have a 50% chance of making it then you almost certainly work in an advertising agency and have missed the point of all this)

 

Coming back to the enigma of what constitutes a "cascade" then remember that only 2% of the tweets created any cascade at all. This does not mean that these 2% created the holy grail of going viral/mainstream.  So I think we should all set our expectation low on how many (marketing) virals actually might go viral and hence how many we need to put out there to deliver what we promise.

 

Perhaps this all come back to what different cultures mean by "we need to put out some WoM ideas" in the sense of how many exactly is "some".

 

I would venture that design agencies typically believe in the one idea, ad agencies are prepared to put out a campaign of up to three executions, PR agencies will punt out 7 or so stories and the digital folk (PPC, SEO, banners) will try dozens of variants to find out what works.

 

Perhaps this reveals the real tension here - should you spend lots of time crafting the perfect sacred flame or chuck out loads of cheap matches and see what takes?

 

So somewhere within all the hype, as the man on Blind Date used to put it, "the choice, as ever is yours"

 

Comments

[ganderson100199]

Well thanks for the info. I learned a few things which can help my cause for Affordable SEO Hampshire.

[PPC Management Companies]

Wow. I don't know if I have enough of a math brain to understand all those numbers. But it does pose the interesting question of a small focused campaign or a broad splash style marketing plan. Definitely something to think about for next time.

[SEO Agency]

I agree with the above post, these are a lot of complicated but interesting numbers. I think word of mouth campaigns only really work if you have a REALLY interesting cause to talk about. Something that affects almost every person on this planet. Otherwise, it's probably wiser to spend your advertising money on more traditional marketing techniques if you wanna see a good ROI. Thanks for sharing!

[Local SEO Services]

Interesting info. Well worth to think about. Especially the little quote from Voltaire in the beginning. But coming from a PR background I would say that the more outlets you use, the more likely it is that someone will find it newsworthy and dig deeper into it. However, don't ever sacrifice quality over quantity in these circumstances.

[Professional Web Design]

I think testimonials activities only really work if you have a REALLY exciting cause to discuss. Something that impacts almost every person in the world. Otherwise, it's probably smarter to invest your promotion money on more conventional promotion models if you want to see a good ROI. Thanks for sharing!

[Business Opportunities]

Nice information but very complicated numbers.......i am very weak in maths so unable to understand properly.....
but sounds very much interesting and useful.

[Maths Confusion]

Great post thanks so much for sharing. I have to say that I'm not taking in all the figures but I'm getting the gist.

Very interesting read..

[jerusalem hostel]

Indeed, Its wonderful information. I must say that I am enjoying reading of your post. You explained everything in very simple & clear language..Thanks a ton :)

[chaussure mbt femme]

Großen Beitrag vielen Dank für das Teilen. Ich muss sagen, dass ich nicht nehmen in allen Figuren, aber ich bekomme das Wesentliche.

[casquette rockstar]

Nizza Informationen aber sehr komplizierte Zahlen ....... Ich bin sehr schwach in Mathematik so nicht richtig verstehen .....

[gorra beisbol]

Fasziniert von seiner guten Blog, danke.

[ralph lauren tee shirt]

Estoy de acuerdo con el post anterior, se trata de un montón de números complejos, pero interesante. muchas gracias!