Marketing Aforethought
Precis: 90% of all high-tech marketing is an afterthought – something that kicks in only once the plan to build a product is already cast into stone. In fact, many high tech companies throw the marketing switch only after the product is already built. This is one of the reasons so many product ideas end up gathering dust in the great Fry's remainder bins of life…
A typical course of events in a tech company looks something like this:
- Someone (often a switched-on geek) gets a cool idea for a new product or new product feature.
- The idea sells internally, the product gets built.
- The company turns around and tries to find someone to buy the new product.
- Someone in management kicks the marketing drones into action: "Oi! Wake up and do some advertising or sumfing, willya?!" (for many high tech companies "advertising" is synonymous with "marketing").
- The marketing dept. realises that they have no idea if a market even exists for the new product, so they have to start down the whole path of "identifying the buyer", "building the message", "presenting the ROI", etc. etc.
- No one buys the darn thing, so they start working on version 2.0
- Lather, Rinse, Repeat...
This kind of behaviour would be absolutely unthinkable in any traditional marketing shop.
Imagine General Motors unveiling a new production car, based entirely on some R&D guy’s idea of "cool". It has three wheels, a joystick, runs on methane and is capable of doing over 300 mph.
Then no one buys it. Because no one bothered to check whether "the market" actually wanted a superfast, drive-by-wire, 3-wheeled roadster.
Of course – this would probably never happen. Bricks & mortar companies like GM still do masses and masses of traditional market research, requirements analysis, product planning, even focus groups (argh) – the full (real) marketing-before-the-fact kind of thing.
In the tech business, however, we all know that the pace of innovation is often the most important driving factor. "Time-to-market" is all – often regardless of whether or not a "market" actually exists. Frame the whole thing in "Internet Time" and the situation gets even worse. Time to market becomes almost moebian – disappearing up our own fundaments long before we’ve even shipped release 1.0.
BTW, I’m convinced this is why so many marketing pitches are so light on “benefits” and confuse feature-pitching with storytelling.
Think about it: how many completely clueless product presentations have you sat through, where Mr. or Ms. shiny product marketing person introduces a slide with: "Now, let’s look at some of the benefits of our solution" -- only to bring up a list of half a dozen tedious me-too technical features of the technology?
Moments like this always reduce me to a frothing, touretting maniac. The calmest response I can muster is to ask: "Yes, very interesting, but so c*#$!*%ing what?"
So your product can flip a zillion quantoflops a nanosecond - what does it do for me?
So you use a peer-to-peer collaboration model – how does that make my life easier?
So you offer XML-based remote administration - why would I need one of those, then?
For years, high-tech companies have been building products that really don't quite do what customers want them to do.
For years, these same companies have been desperately trying to market their almost-but-not-quite-useful ideas on the basis of retrospective identification of some perceived "need" in their target market. Build a product, then see if you can find someone who needs it.
Large sections of the industry, numerous “ground-breaking” products, entire technology categories have sprung forth fully formed out of a vacuum of zero market contact.
Want proof? Think of Lotus Notes. I mean: what a pile of pants. Really.
Pretty much the entire product plan for this thing was: “1-2-3 got us to the top of the dogpile, now build me another killer app”.
Magellan hadn’t done it for them (shame – cool product, ahead of its time, looking to fit a market need that didn’t quite exist yet). They needed a new strong revenue driver.
From what I recall of the whole kickstart (I attended the UK launch), when Notes first shipped, no one (Lotus included) had the faintest idea what the hell it really was and what in the name of arse anyone might need it for.
It was a perfect example of a "cool" idea that didn’t actually fit any existing need.
Not quite a 4GL, sort of like email, not quite workflow, kinda sortofa DMS. Fat, buggy, counter-intuitive and defying most by-then-accepted UI standards (like: right click to close a window, fercrissakes - I ask you).
But they got lucky. Almost accidentally, it seems, a few people at some of the top sys integrators and Big 6 (as there still were, back then) realised there was a huge services opportunity helping Global 2000 companies figure out what the feck Lotus Notes was all about.
E&Y and others signed big noisy deals for shedloads of Notes licences (their logic: “if we throw enough copies at our consultants and force them to start playing with it, maybe some of them will work out how we can make money out of it).
The rest is history.
Customers even got swept up in the giddy exhilaration of it all: signing enterprise license deals right in there alongside the sales guys, high-fiveing like crazy, pumped and energized by the sheer mad momentum of it all. Only to come abruptly to their senses several months into the implementation, when they realised they’d blown millions and squandered countless man hours building something nobody wants, no one can use, no one can support, no one even likes.
The Standish Group did some research into large scale software implementations a while back. Scary stuff: “31.1% of projects will be canceled before they ever get completed… 52.7% of projects will cost 189% of their original estimates."
What’s wrong with this picture?
Oh, and please – don’t start me on the whole “solutions” thing. Next time someone describes their latest widget as a “solution”, demand that they show you the problem.
In six words: What. Specifically. Does. Your. “Solution.” Solve?
Just try saying this to a software sales guy without tacking the thought "you greasy wanker" onto the end.
“Solutions”, in the mouths of the many-too-many hucksters, is a completely devalued word that simply means: “another clueless product with a nod in the direction of over-priced services, maybe wrapped up in an implementation methodology, if you’re really lucky.” Give me strength.
I’ve been in the meetings where these things get dreamed up:
"People aren’t buying enough of our product.
"We’ve done all the same things as our competitors: loaded the product with pointless features, made sure that it bloats out once installed to consume all available resources, introduced enough bugs to keep the maintenance revenue rolling for years – but still we’re not shipping enough.
"People can’t figure out why they need our product or work out how to justify the expense to the higher-ups.
"Ergo: we need a professional services team and a bunch of implementation partners to build solutions".
It’s at this point that someone inevitably mentions Geoffrey Moore and the whole thing really starts to go pear-shaped.
And oh yeah – ”software maintenance”, that’s another good one. Incompetence tax.
“Our stuff’s rubbish, please pay us two year’s cretin insurance, in advance, so we can hire some decent people to fix it.”
OK, calm down. So, there are exceptions, of course.
Some software products are still designed, from scratch, with a reasonable idea of what the market wants or what gaps in other products exist that REALLY need to be filled. These are often the products built by the users, with the users, for the users.
Some products even work. Really well. Many of the best products on the market have come directly from engineers with no market input.
The problem is it takes a ratio of about 1500:1 to get that single great idea.
The remaining 1499 are just so shamefully, woefully bad, the industry should stand in the corner with its head down for the next 30 years.
OK. So. Lot’s of energetic invective – but do I have any answers?
Going back to The Standish Group thing, the most telling section in their research was the analysis of why software projects so often fail. The top two failure factors were: “lack of user input” and “incomplete requirements & specifications”.
In other words, a lot of software is junk simply because it doesn’t do what people want and the people developing it are kept too far away from any ability to understand what they should be building.
In the last two years we’ve seen entire companies built like this -- conceived with no realistic understanding of specific user requirements and little connection to what the market actually wants or needs.
The stem of the problem is simply the lack of involvement of customers in the early stages of requirements gathering & system planning. This is real Product Marketing 101, for goodness sake.
Requirements, baby – it’s all about requirements. Here's a clue (sit up there at the back, this is the money shot):
If something is not required, don’t build it. If you know what users require – build that.
“If you build it, they will come” simply doesn’t work in the real world.
Closer to the truth, in many cases, is: “If you build it, no one will show up unless they really need it (even if it’s funky as all heck)”.
I could go on in this vein for many pages more, but even I’m starting to tire of the rather shrill tone of this piece.
Just please, do me a favour: next time you get a great idea to build a new product – stop.
Go out into the world first. Find some people who do something for a living beyond “use software”. Talk to them. Ask them what they actually need to make their specific jobs easier.
In the old days, this would have been called “market research”.
That is: not collecting analyst numbers on the size of the “market” in revenue or seat counts, or shifting yerselves ever north-eastwards through some vapid magic quadrant, but rather actual research into whether or not a market exists for an idea. It may seem like a great idea, but does anyone actually need it? Apply some marketing aforethought.
If you find something people want, by all means build it. Build it well, and keep checking that your engineers and customers are communicating without too many translation layers. You need to reality check user requirements of course, but you also need to make sure that genuine human requirements can actually make it into the products you create.
Then we might actually start to see things that resemble “solutions” to some of the problems we've created for ourselves.
A typical course of events in a tech company looks something like this:
- Someone (often a switched-on geek) gets a cool idea for a new product or new product feature.
- The idea sells internally, the product gets built.
- The company turns around and tries to find someone to buy the new product.
- Someone in management kicks the marketing drones into action: "Oi! Wake up and do some advertising or sumfing, willya?!" (for many high tech companies "advertising" is synonymous with "marketing").
- The marketing dept. realises that they have no idea if a market even exists for the new product, so they have to start down the whole path of "identifying the buyer", "building the message", "presenting the ROI", etc. etc.
- No one buys the darn thing, so they start working on version 2.0
- Lather, Rinse, Repeat...
This kind of behaviour would be absolutely unthinkable in any traditional marketing shop.
Imagine General Motors unveiling a new production car, based entirely on some R&D guy’s idea of "cool". It has three wheels, a joystick, runs on methane and is capable of doing over 300 mph.
Then no one buys it. Because no one bothered to check whether "the market" actually wanted a superfast, drive-by-wire, 3-wheeled roadster.
Of course – this would probably never happen. Bricks & mortar companies like GM still do masses and masses of traditional market research, requirements analysis, product planning, even focus groups (argh) – the full (real) marketing-before-the-fact kind of thing.
In the tech business, however, we all know that the pace of innovation is often the most important driving factor. "Time-to-market" is all – often regardless of whether or not a "market" actually exists. Frame the whole thing in "Internet Time" and the situation gets even worse. Time to market becomes almost moebian – disappearing up our own fundaments long before we’ve even shipped release 1.0.
BTW, I’m convinced this is why so many marketing pitches are so light on “benefits” and confuse feature-pitching with storytelling.
Think about it: how many completely clueless product presentations have you sat through, where Mr. or Ms. shiny product marketing person introduces a slide with: "Now, let’s look at some of the benefits of our solution" -- only to bring up a list of half a dozen tedious me-too technical features of the technology?
Moments like this always reduce me to a frothing, touretting maniac. The calmest response I can muster is to ask: "Yes, very interesting, but so c*#$!*%ing what?"
So your product can flip a zillion quantoflops a nanosecond - what does it do for me?
So you use a peer-to-peer collaboration model – how does that make my life easier?
So you offer XML-based remote administration - why would I need one of those, then?
For years, high-tech companies have been building products that really don't quite do what customers want them to do.
For years, these same companies have been desperately trying to market their almost-but-not-quite-useful ideas on the basis of retrospective identification of some perceived "need" in their target market. Build a product, then see if you can find someone who needs it.
Large sections of the industry, numerous “ground-breaking” products, entire technology categories have sprung forth fully formed out of a vacuum of zero market contact.
Want proof? Think of Lotus Notes. I mean: what a pile of pants. Really.
Pretty much the entire product plan for this thing was: “1-2-3 got us to the top of the dogpile, now build me another killer app”.
Magellan hadn’t done it for them (shame – cool product, ahead of its time, looking to fit a market need that didn’t quite exist yet). They needed a new strong revenue driver.
From what I recall of the whole kickstart (I attended the UK launch), when Notes first shipped, no one (Lotus included) had the faintest idea what the hell it really was and what in the name of arse anyone might need it for.
It was a perfect example of a "cool" idea that didn’t actually fit any existing need.
Not quite a 4GL, sort of like email, not quite workflow, kinda sortofa DMS. Fat, buggy, counter-intuitive and defying most by-then-accepted UI standards (like: right click to close a window, fercrissakes - I ask you).
But they got lucky. Almost accidentally, it seems, a few people at some of the top sys integrators and Big 6 (as there still were, back then) realised there was a huge services opportunity helping Global 2000 companies figure out what the feck Lotus Notes was all about.
E&Y and others signed big noisy deals for shedloads of Notes licences (their logic: “if we throw enough copies at our consultants and force them to start playing with it, maybe some of them will work out how we can make money out of it).
The rest is history.
Customers even got swept up in the giddy exhilaration of it all: signing enterprise license deals right in there alongside the sales guys, high-fiveing like crazy, pumped and energized by the sheer mad momentum of it all. Only to come abruptly to their senses several months into the implementation, when they realised they’d blown millions and squandered countless man hours building something nobody wants, no one can use, no one can support, no one even likes.
The Standish Group did some research into large scale software implementations a while back. Scary stuff: “31.1% of projects will be canceled before they ever get completed… 52.7% of projects will cost 189% of their original estimates."
What’s wrong with this picture?
Oh, and please – don’t start me on the whole “solutions” thing. Next time someone describes their latest widget as a “solution”, demand that they show you the problem.
In six words: What. Specifically. Does. Your. “Solution.” Solve?
Just try saying this to a software sales guy without tacking the thought "you greasy wanker" onto the end.
“Solutions”, in the mouths of the many-too-many hucksters, is a completely devalued word that simply means: “another clueless product with a nod in the direction of over-priced services, maybe wrapped up in an implementation methodology, if you’re really lucky.” Give me strength.
I’ve been in the meetings where these things get dreamed up:
"People aren’t buying enough of our product.
"We’ve done all the same things as our competitors: loaded the product with pointless features, made sure that it bloats out once installed to consume all available resources, introduced enough bugs to keep the maintenance revenue rolling for years – but still we’re not shipping enough.
"People can’t figure out why they need our product or work out how to justify the expense to the higher-ups.
"Ergo: we need a professional services team and a bunch of implementation partners to build solutions".
It’s at this point that someone inevitably mentions Geoffrey Moore and the whole thing really starts to go pear-shaped.
And oh yeah – ”software maintenance”, that’s another good one. Incompetence tax.
“Our stuff’s rubbish, please pay us two year’s cretin insurance, in advance, so we can hire some decent people to fix it.”
OK, calm down. So, there are exceptions, of course.
Some software products are still designed, from scratch, with a reasonable idea of what the market wants or what gaps in other products exist that REALLY need to be filled. These are often the products built by the users, with the users, for the users.
Some products even work. Really well. Many of the best products on the market have come directly from engineers with no market input.
The problem is it takes a ratio of about 1500:1 to get that single great idea.
The remaining 1499 are just so shamefully, woefully bad, the industry should stand in the corner with its head down for the next 30 years.
OK. So. Lot’s of energetic invective – but do I have any answers?
Going back to The Standish Group thing, the most telling section in their research was the analysis of why software projects so often fail. The top two failure factors were: “lack of user input” and “incomplete requirements & specifications”.
In other words, a lot of software is junk simply because it doesn’t do what people want and the people developing it are kept too far away from any ability to understand what they should be building.
In the last two years we’ve seen entire companies built like this -- conceived with no realistic understanding of specific user requirements and little connection to what the market actually wants or needs.
The stem of the problem is simply the lack of involvement of customers in the early stages of requirements gathering & system planning. This is real Product Marketing 101, for goodness sake.
Requirements, baby – it’s all about requirements. Here's a clue (sit up there at the back, this is the money shot):
If something is not required, don’t build it. If you know what users require – build that.
“If you build it, they will come” simply doesn’t work in the real world.
Closer to the truth, in many cases, is: “If you build it, no one will show up unless they really need it (even if it’s funky as all heck)”.
I could go on in this vein for many pages more, but even I’m starting to tire of the rather shrill tone of this piece.
Just please, do me a favour: next time you get a great idea to build a new product – stop.
Go out into the world first. Find some people who do something for a living beyond “use software”. Talk to them. Ask them what they actually need to make their specific jobs easier.
In the old days, this would have been called “market research”.
That is: not collecting analyst numbers on the size of the “market” in revenue or seat counts, or shifting yerselves ever north-eastwards through some vapid magic quadrant, but rather actual research into whether or not a market exists for an idea. It may seem like a great idea, but does anyone actually need it? Apply some marketing aforethought.
If you find something people want, by all means build it. Build it well, and keep checking that your engineers and customers are communicating without too many translation layers. You need to reality check user requirements of course, but you also need to make sure that genuine human requirements can actually make it into the products you create.
Then we might actually start to see things that resemble “solutions” to some of the problems we've created for ourselves.
