This rather violates the idea of 'sell your by-products' though doesn't it?
I agree for the initial phase of a startup this is a good strategy, but does it really make sense to ignore additional revenue streams which might actually be better than those you originally decided to concentrate on?
I think the point is that those byproducts are just that — byproducts, not part of your core business plan. If your business is relying on byproducts rather than just exploiting them when they appear, you have kind of a fragile model.
If the other revenue stream is better, then switch to it. But the point of the article is, no, don't sell your by-products. Focus on your best revenue stream.
> If you are in a position to say: "No problem, we didn’t really need them anyway," then why were you wasting time talking to them at all?
Maybe because it was a good way (or a potentially good way) of making money?
Maybe because you wanted to have a foot in the door in what might turn into a full pivot?
The example of Google is a terrible one, because sure they make all their money from advertising, but based on the "You want to focus on the most promising possibility, exclusively, until it’s no longer the most promising possibility" logic, Google shouldn't be doing anything except working on their search engine, not branching off and spending time/money on gmail, maps, plus, etc. etc. etc.
Perhaps its just my read, but I think the original post confuses revenue streams with business models with products. It seems to flip from business model to product design to revenue streams back to business models and finishes with a vague statement about "core".
Selling complimentary products and services to overlapping and peripheral market segments is an extremely valid and sustainable strategy (Apple). Using different methods of distribution to attack different market segments with similar products works extremely well (Amazon, Google). Launching additional lines of business to attack unrelated markets using a singular strong brand also has merit (Virgin).
The key really is "don't have a shitty business model" not "don't have multiple revenue streams".
On the other hand, I really appreciate Rob's points about how clever often means convoluted (I am always wary of "complicated") and that mediocrity and a lack of focus can be dilutive.
One wonders what a company like 37 Signals would be classified as.
Are they one business model? Cause they certainly have several different successful products, although i'd claim that they have a single methodology that's applied all of them (hosted web apps that don't require humans to scale up).
Likewise Microsoft... OS manufacturer? Office Suite product company? Video game manufacturer? Web-based something or other?
I think the quote from Alex was in the context of a hypothetical business model in a pre-traction startup - it was a tangible way to make the point about the problem with convoluted models at that stage.
I agree with other commenters, that sometimes you try other revenue streams in order to see how the market or potential customers might respond. If you see any good signs then one may choose to focus and strengthen these options.
This is ludicrous. The fact that Google only has one revenue stream is one of their greatest weaknesses, and they know it, which is why they're aggressively expanding into everything under the sun. And as long as we're cherry-picking examples, why not Microsoft or Apple or GE?
Ultimately, using Google or any other Fortune 500 company as a model for how you should do business as a fledgling startup is probably not a good idea.
It is not the best written article but I think that some people here are wilfuly misrepresenting it (maybe being a bit defensive?).
He is clearly talking about clarifying your business model in the early (pre-launch?) days of a startup. In that context everything he says about simplicity and focus should be kind of self-evident.
To crudely paraphrase, if you haven't got one solid good idea (revenue stream, whatever) then 5 bad ones probably won't help you.
Agree with some that the post is a bit confusing. I think the point is, Pick one: transaction fees, ad sales, professional services, software licensing, subscription fees, etc.
Rob focused on the aspect of Alex' interview about the benefits of simple models to test. This helps invalidate them quickly, enabling you to progress. The flip-side is that you keep a portfolio of business model hypotheses, which you can test sequentially.
So if you put all 5 revenue sources in a single model, and one of them fails, you can justify continuing testing that business model to yourself because there's always another revenue source to test. (This is compounded quickly by all the possible permutations of value proposition, customer segment, channel, relationship type and engines of growth.)
Whereas, if you have simple models to test, and the single revenue source fails, you can take action on this by crossing off that specific model and moving on to the next one.
Simpler models in this case help you pull out discrete and actionable learning, and to focus on one thing at a time, and move on quickly if the concept doesn't check out.
17 comments
[ 0.13 ms ] story [ 44.7 ms ] threadI agree for the initial phase of a startup this is a good strategy, but does it really make sense to ignore additional revenue streams which might actually be better than those you originally decided to concentrate on?
You cannot begin with a business plan/model scattered all over the place, with some core product and several byproducts.
Maybe because it was a good way (or a potentially good way) of making money?
Maybe because you wanted to have a foot in the door in what might turn into a full pivot?
The example of Google is a terrible one, because sure they make all their money from advertising, but based on the "You want to focus on the most promising possibility, exclusively, until it’s no longer the most promising possibility" logic, Google shouldn't be doing anything except working on their search engine, not branching off and spending time/money on gmail, maps, plus, etc. etc. etc.
Selling complimentary products and services to overlapping and peripheral market segments is an extremely valid and sustainable strategy (Apple). Using different methods of distribution to attack different market segments with similar products works extremely well (Amazon, Google). Launching additional lines of business to attack unrelated markets using a singular strong brand also has merit (Virgin).
The key really is "don't have a shitty business model" not "don't have multiple revenue streams".
On the other hand, I really appreciate Rob's points about how clever often means convoluted (I am always wary of "complicated") and that mediocrity and a lack of focus can be dilutive.
Are they one business model? Cause they certainly have several different successful products, although i'd claim that they have a single methodology that's applied all of them (hosted web apps that don't require humans to scale up).
Likewise Microsoft... OS manufacturer? Office Suite product company? Video game manufacturer? Web-based something or other?
Ultimately, using Google or any other Fortune 500 company as a model for how you should do business as a fledgling startup is probably not a good idea.
My startup-land rebuttal for this claim would be 37signals: http://37signals.com/svn/posts/1123-theres-more-than-one-way...
Early on, as a developing firm, Google focused on that single revenue stream to great success.
He is clearly talking about clarifying your business model in the early (pre-launch?) days of a startup. In that context everything he says about simplicity and focus should be kind of self-evident.
To crudely paraphrase, if you haven't got one solid good idea (revenue stream, whatever) then 5 bad ones probably won't help you.
So if you put all 5 revenue sources in a single model, and one of them fails, you can justify continuing testing that business model to yourself because there's always another revenue source to test. (This is compounded quickly by all the possible permutations of value proposition, customer segment, channel, relationship type and engines of growth.)
Whereas, if you have simple models to test, and the single revenue source fails, you can take action on this by crossing off that specific model and moving on to the next one.
Simpler models in this case help you pull out discrete and actionable learning, and to focus on one thing at a time, and move on quickly if the concept doesn't check out.