I don't know if this is quite public knowledge yet, so: human-assisted onboarding is sort of catching on at a lot of SaaS companies in the B2B space, at price points which are a bit north of the top end of typical 37signals/etc apps. (Squarely in the zone for this new product, though.)
A lot of people find it very, very difficult to get started from a blank slate. Automated onboarding/email sequences/video tutorials/etc gets you to a point, but it still leaves guesswork in the process and (candidly) one can always blow off a computer but blowing off a person who has been talking to you for weeks is socially harder to do.
This also allows you to do basically a hybridized product/consulting model, since if (without loss of generality) Jason Fried is spending 2 hours for you talking about your business' challenges as he gets you up to speed on KYC, that's essentially equivalent from the customer's perspective to "Rent Jason's brains", but (crucially) the majority of the economic activity actually comes from the product whose use is enabled by the mini-consult rather than through consulting itself. This lets it scale to the moon. (Also, as every consulting company at scale will tell you, you can have Jason Fried's insights on your company actually delivered by an understudy. 37signals certainly doesn't lack for smart people who understand the 37signals playbook.)
This is currently giving me fits, since I haven't figured out a great way to deliver it consistently for non-enterprise accounts given that I'm 14 time zones away from many customers, but it's one of the big things I'm working on currently. The word on the B2B SaaS grapevine is it works very, very, very well at moving $X,000 to $X0,000 per year SaaS accounts.
I'll blog about this if I ever manage to systemize it successfully. (As usual, the plan is 1) Throw stuff at the wall and figure out what works. 2) Write down what worked and execute against that plan. 3) Pay someone to do #2. I'm currently somewhere in stage #1.)
Hey Patrick--as the person who (may have) introduced you to this, I should share what we're doing. We are just about to hire another full-time person to help us with this as we scale out our consulting-as-onboarding process. I promise to write a blog post once I have more data on how well this works, but I will say this: You can get a well-qualified person in the U.S. who's smart, technically savvy (not a programmer but knows his/her way around computers), and has a bachelor's degree in a related field for $40K/year in salary--and that's a step up from where many of these folks are today.
I'm really excited to have a FT person working on this, and I think in the future we'll see the "hacker/hustler/designer" trio that Dave McClure et.al. are so fond of speaking of expand into four people, with the fourth being a customer service/customer onboarding role. In fact, this is our fourth FT person on the team (other 3 are two savvy developers plus me, the business brain.) A designer will be our fifth hire.
More data: We plan to implement consulting-as-onboarding for any plan priced over $499/mo (which is our sweet spot right now), and, in fact, "do it for them" in terms of getting all their data into our system (which is where many of our customers are struggling now.) If it works and pays off in terms of customer acquisition and retention, I have full plans to do this for EVERY customer. I learned how to do this in the hosting industry--Rackspace was my example 10 years ago, and I did this with massive success with my hosting company as well. It was also one of my top tips for WP Engine when I consulted there.
P.S. Patrick, if you're still reading, we talked about doing a podcast, and this is what we should talk about. I'm game--I'll shoot you an email.
I have my reservations about this, because it feels like B2B SaaS is now going upmarket by reaching for the familiar Enterprise, high-touch sales process.
We've all been through this, and we know where it leads.
You have to bring on dedicated high-touch sales people to "scale" the business.
Going up market means compromising more on the scope of what the products do. As individual deals become more lucrative, the incentive to close them grows. As this incentive increases, the products often suffer and you get the Bananas in the Lasagna problem (See Jason Fried: http://entrepreneursunpluggd.com/blog/37signals-jason-fried-...).
So my question to the folks advocating this "new" approach is, how are you going to avoid all those things?
You have to bring on dedicated high-touch sales people to "scale" the business.
So it's very much different than a traditional sales guy in the Oracle sense. Some companies brand them as Customer Success Advocates or "coaches" or something similar. You aren't asking for "the sale" per se -- the customer is generally pretty onboard that they want it and will often have given their credit card already. You're just making sure that they actually get to using it successfully. This involves handholding, counseling, motivating, and being organized and persistent rather than bashing down objections, consultative sales, and always-be-closing.
In addition to it not reading like "traditional sales" from either side of the telephone, the background/inclinations/etc of the employees who do this are quite different. One obvious example: at many companies, these folks tend to run towards ladies, much like e.g. marketing teams or customer support and in contrast to how engineering teams or sales teams in B2B companies rarely tend to have large concentrations of ladies. (Positive claim not a normative claim, guys: regardless of the optimal state of the world, it is an objectively true statement that the average number of Y chromosomes per team member varies widely across departments in B2B software companies.)
As to avoiding it corrupting the flavor of the company: I'm pretty sure 37signals has a corporate culture which isn't going to turn into Oracle anytime soon, and it's unlikely that this would radically change the character of my business either. (Money generally doesn't affect people nearly as much as people expect it to.)
As I mentioned in my post, high-touch, full-service doesn't have to be the permanent model. Our plan isn't to do it this way forever. Our plan is to do it this way for a while so we can learn everything we can. Then later on we can automate this and make it self-service. The product will be a lot better off because of it. This was pg's point in his original post as well.
As a data point from another SaaS company owner, we thought this would be temporary too, but we found out it was so valuable in increasing customer retention and engagement that we are now planning to make it permanent (see my post above--we're hiring a FT person for this now.) The cost-effectiveness of this is off the charts. Patrick found out about this from me when I spoke about it at Microconf and I already have people from there telling me it's made them $500/hr+ (I believe it, as I've seen similar numbers in my own SaaS company.)
Oh, I should add: If you want people to rave about your company, there's nothing better than this onboarding process to do so. People become sort of religious about your business after a first-touch customer experience like this one. Probably preaching to the choir if you are Jason Fried, but this goes double for everyone else reading this who's building a software company.
I'm glad to hear that companies such as yours that are doing this are not necessarily considering making it permanent. I read PG's original article, and although you make reference to it, I felt he was addressing a different sort of issue than it sounds like you are. In your case, you actually went out of your way to engineer internal systems that let you do things manually, and you probably could have spent the same time and engineering resources you used to do that to instead develop the automation you don't have yet.
I may be in the minority, but let me give you some perspective as a potential customer who has experienced this kind of hands-on "service" from other SaaS-y outfits: I WANT automation from your service. I have worked with 2 companies that made me jump through all kinds of hoops to get an account set up with them when I could have had a usable account provisioned and active in 5 minutes had they automated their processes. As a result, it took days, because (surprise) their employees have their hands full, and I also have better things to do (you know, actual work responsibilities and customers of my own to respond to) than to have my hands held jumping through all of your contrived and artificial hoops just to use your service. That's just a waste of my time and yours and something that only ends up causing unnecessary delays to service turn-up, and in both cases, it seriously soured my taste for the service and almost made me contemplate throwing in the towel and going with a different solution.
This isn't the feeling that you want to leave your new customers with who are trying very hard to give you their money. You mean every time we need to make a change to service parameters, I need to submit a spreadsheet with my change request to somebody that's going to plug it in on your side by hand? MAYBE 24 hours later, if I'm lucky? Forget that: I'll go with somebody that gives me an API or a web portal that I can use to make the changes myself in a matter of minutes.
But what about "post-sale"? I think there will always be customer demand for concierge-level of service. Some SaaS models do lend themselves to self-service, but others don't. Knowing that distinction is key.
Anyone who's signed up at the concierge-level of service will always get that level of service. And if we ultimately go self-service, I'd still keep the concierge-level of service around, just at a premium. We might even consider adding a concierge-level of service tier to Basecamp.
Watch Gail Goodman's presentation "How to negotiate the long, slow, SaaS ramp of death" [1] for an interesting discussion of this. (Skip ahead to 20:42 for her description of "coaches".)
Is this a new writing style I'm unaware of or just plain laziness?
Get to the point blog writers!
Make your assertion clearly and right at the top, explain it if it could be misconstrued so that we know what the hell is being discussed.
Then go on to write about it.
It seems like every second blog post linked to here is a long rambling missive that only reveals the point somewhere minutes into reading it long after my eyes have glazed over.
Yes, assume your reader has just woken up and has 20+ tabs open they're trying to speed-read through before they finish their bagel and coffee and rush off to work. They want the firehose, not Dostoyevsky.
Good rule of thumb for this kind of audience - invert your essay structure. Put the concluding paragraph first, then work backwards through the supporting paragraphs till you get to the original question.
That style is much more engaging than traditional, overused, boring form of "Intro para -> 3x supporting para -> conclusion". Instead make a strong/counterintuitive/nonobvious/etc unsupported assertion right off the bat, then work backwards to support it.
That initial assertion captures peoples attention much more effectively than a slow build up, b/c they then want to read on to find out if the author can actually support it, and are gratified when you do.
> Good rule of thumb for this kind of audience - invert your essay structure. Put the concluding paragraph first, then work backwards through the supporting paragraphs till you get to the original question.
> That style is much more engaging than traditional, overused, boring form of "Intro para -> 3x supporting para -> conclusion".
No, its not "more engaging" than the traditional 5 paragraph essay, because in the traditional 5-paragraph essay format, the first sentence (the thesis statement) is the unsupported (to that point!) position which the rest of the essay supports. You aren't arguing for "inverting" the "traditional, overused, boring" 5 paragraph essay format, you are arguing for using the "traditional, overused, boring" 5 paragraph essay format in exactly the manner it is traditionally taught.
> Haha, fine, then I'm arguing for inverting whatever the current standard seems to be
Insofar as long, rambling intros that don't get quickly to a point are the norm, that's a standard only insofar as "people don't bother to apply any writing advice anyone has ever given" is a standard.
What I described isn't a TLDR, it's still a full essay, just with the most powerful/engaging content up front, instead of buried at the end.
In fact I've recently found myself reading blog posts backward just to skip the rambling pre-amble and get straight to the meat of it, but I still backtrack from the end to the beginning, filling in the details.
I agree with you, except that beginning a piece with an argument and then following that argument with supporting evidence isn't inverted -- it's Essay Writing 101.
That's not quite what I meant. I'm saying, start with the conclusion to the argument, then work backwards through the supporting evidence to the argument, and conclude by reviewing how the originally stated conclusion satisfies it.
Can't think of a good example atm, but will try to find one and have it handy in case this topic ever comes up again.
This is true - yet I'm pleased to see that the 37signals blog posts which are starting make it to the top of HN are those with more depth. For a while, it seemed like all of the ones making the front page were simple platitudes with no concrete examples or solutions e.g. http://37signals.com/svn/posts/3513-when-empathy-becomes-ins... .
Interesting...there was an article on HN recently that posited a major reason for the 'failure' of the Facebook Platform was an over-engineered attempt to mitigate spam and promote quality control through automation...instead of curating apps by hand, a la the iOS App Store.
pg's original article was great, and it's nice that 37signals are along for the ride - but let's be clear, Jason's article is about "artisanal high-touch", not the real deal.
The real-deal looks like email lists, cron jobs, Google Docs, throwaway Ruby scripts, Access databases. It's operational and technical debt that's the least worst option. Artisanal high-touch has a bevelled drop-shadowed button saying "They paid us! Enter the transaction manually..." and full Campfire integration.
Well, pg gave several examples of "things that don't scale" in his article, and I believe Jason is referring to focusing on individual users, not doing tasks manually. He mentions eschewing automation but he's talking about automation of client interations, not automation of all the software processes.
In terms of software I agree though, clicking pretty boxes is just a different type of automation. I am co-opting the phrase "artisanal high-touch."
Otherwise known as professional services. I think that's what some folks miss when they think about SAAS - it's going to be about service more than anything else despite the network delivery.
1. Know Your Company is, at minimum, a $2500 service. The amount of time you can spend with a given customer is huge, in that respect, because the customer support you're doing for that one customer is proportionally much more valuable. If I have a SaaS offering billing out at $20/mo., how much time can I feasibly spend with a customer (of which there are multitudes) before I have to think of ways to avoid timesinks?
2. The "manually adding people to a database" thing seems contrived. It's valuable because you learn the departments and the employees: so what's wrong with automating the excel/email uploads and having the script spit out an org chart and dossier to the printer? (That isn't a rhetorical question -- wouldn't a birds-eye view of an organization be more helpful?)
(Overall, I absolutely agree with Jason's point! But I think the argument is a bit more nuanced than the way he presents it.)
1. This is part of the point. Maybe $20/month isn't the right model. Maybe there's a different model you can consider where you'd have fewer customers, but those customers would be at a much higher price point. Who knows. Point is, it's your call and it's all possible depending on how you approach it.
2. Importing the spreadsheet of customer names will likely be one of the first things we automate. It's less insightful than the other things on the list, but I still believe doing it manually is teaching us something right now.
"Maybe $20/month isn't the right model. Maybe there's a different model you can consider where you'd have fewer customers, but those customers would be at a much higher price point."
Another model to consider would be to have different private label companies (that you control) serving up essentially the same product, priced differently, and offered through a different company name.
For example you have basecamp where you can manage unlimited projects at $3000 per year.
But it's possible that Boeing might pay $50,000 per year if they got even a higher level of service and a personal account manager. [1]
Now you could do this by adding a tier to basecamp but you could also be your own competition with basecamp just for the bespoke offering.
Figures arbitrary to make a point.
[1] Sold by a professional sales force or some other higher cost method than adwords.
For example I'm helping to investigate, as a favor, solutions for Physician scheduling. I'm amazed at how cheap the service is. I know that the group that wants this would easily pay much more to solve the problem in a turnkey fashion. But the current offerings are being sold on price only as a way to get attention. (And not doing a good job at it for that matter.)
I know the article is referring to b2b SaaS products, but we employed a version of this idea on Http://storifi.it
We already knew that people wouldn't know what the game was about, so we sat with each user individually and talked them through the onboard process. We learned alot of things we would not have otherwise spotted and then capitalised on that learning. The web app was just an mvp, but we incorporated all the learning into our mobile app which resulted in a 100% activation rate.
Next step for us is now to finish off the app and concentrate on user engagement for which we'll be following a similar philosophy.
A lot of people find it very, very difficult to get started from a blank slate. Automated onboarding/email sequences/video tutorials/etc gets you to a point, but it still leaves guesswork in the process and (candidly) one can always blow off a computer but blowing off a person who has been talking to you for weeks is socially harder to do.
This also allows you to do basically a hybridized product/consulting model, since if (without loss of generality) Jason Fried is spending 2 hours for you talking about your business' challenges as he gets you up to speed on KYC, that's essentially equivalent from the customer's perspective to "Rent Jason's brains", but (crucially) the majority of the economic activity actually comes from the product whose use is enabled by the mini-consult rather than through consulting itself. This lets it scale to the moon. (Also, as every consulting company at scale will tell you, you can have Jason Fried's insights on your company actually delivered by an understudy. 37signals certainly doesn't lack for smart people who understand the 37signals playbook.)
This is currently giving me fits, since I haven't figured out a great way to deliver it consistently for non-enterprise accounts given that I'm 14 time zones away from many customers, but it's one of the big things I'm working on currently. The word on the B2B SaaS grapevine is it works very, very, very well at moving $X,000 to $X0,000 per year SaaS accounts.
I'll blog about this if I ever manage to systemize it successfully. (As usual, the plan is 1) Throw stuff at the wall and figure out what works. 2) Write down what worked and execute against that plan. 3) Pay someone to do #2. I'm currently somewhere in stage #1.)