5 ms·

Against all of the advice in the world, I'm trying to set things up to avoid basically everything described in the article. It's tricky because the truth is tha

Against all of the advice in the world, I'm trying to set things up to avoid basically everything described in the article. It's tricky because the truth is that a lot of the tactics work and that is why companies implement them. A lot of investors also consider that process to be best-in-class and look for it as a sign of maturity. The counter approach is: - Make complete information available as transparently as possible and don't gate it. - Be forthcoming about weaknesses. Don't force prospects and customers to find them. - Ensure that when a prospect or customer does want to talk to someone, they immediately reach someone who can handle the problem or answer the question (no need for escalations.) - Never have an AI agent call someone unless the customer specifically requests that and be sure that all AI agents immediately disclose that they are AI. - Offer flexible e-mail list subscription options (monthly, quarterly, annually, only release notes, etc.) - If the product is not a fit, try to offer something useful anyway such as a suggestion of another product that might be a better match for their needs. Value based pricing is one of the reasons why a lot of companies end up in these situations. Rather than setting a standard price, the company does a detailed investigation of the customer to try to find out how much value they will gain from using the product and then they set the price based on that determination. Although it maximizes revenue in theory, it is slow and invasive.

portaouflop 2y ago
More power to you but in my experience - people don’t read your documentation - if they find one edge-case that is not covered by your product they concentrate all their energy on it even though the competitor product doesn’t even work for the happy path (but ofc they won’t learn that until too late) - you will have a person on the free tier who is not able to the simplest programming take up all the time of your technical support if you let them The rest (disclose usage of AI, offer multiple newsletters (imo unnecessary work), be upfront if it’s not a good fit) is very sensible and works in my experience. The first three will break your product - but maybe I’m just cynical and was holding it wrong all this time edit: I’m assuming you want to build some kind of investor fuelled unicorn - your approach probably will work for a single person slow but steady growth while you have other side income
- efitz 2y ago
  You need docs. Most users won’t read them. Do a good job anyway because they can save you a TON of support costs if they’re good, even for the few people who do read them. The cost of writing docs is easily offset by even a modest reduction in human support cost, unless you just half-ass your support with LLMs and FAQs and chat with overseas wage slaves. ALSO: your UX needs to warn people about sharp edges and one-way doors, BEFORE your user commits to an action they might regret. These are not incompatible.
  - johnnyanmac 2y ago
    If nothing else, the "smart cows" will read the docs and may even help out on support forums when some of the user base doesn't and decides to complain. Docs are good for everyone and will save time internally and externally. >your UX needs to warn people about sharp edges and one-way doors, BEFORE your user commits to an action they might regret. These are not incompatible. Assuming trump doesn't nix it, the recent FTC ruling should fix this by itself. it should be as easy to unsubscribe as it is to subscribe.
    - eigilsagafos 2y ago
      I don’t think easy unsubscribe falls under «sharp edges and one-way doors». But high friction unsubscribe probably goes hand in hand with sharp edges and one-way doors at those lead you to want to unsubscribe…therefore they make it hard.
  - freefaler 2y ago
    They also are great for the support team, to "educate" where they can find additional info and for linking as help links if the interface needs this.
  - babyshake 2y ago
    Also, it's pretty easy at this point to setup an AI support agent that can reference your docs, and it is a helpful exercise for organizing your own thoughts and being able to communicate them back to yourself. There's not really a good reason to not have good docs. I just assume that if a product is gated behind "let's hope on a call!" it's still vaporware and there's a 50/50 chance that they'll pivot within a few months.
rkagerer 2y ago
Good advice. I created and sold software (and services) for over a decade, and could never have lived with myself if I'd treated customers the way the original article describes.
- johnnyanmac 2y ago
  sadly, they do it because it works, and (not trying to be too offensive): lotta people are dumb. If you're confused, a nice "team mate" will be more comforting than a cold hard doc and web form. Even if they are upcharging you on a bunch of stuff you will never use. They call multiple times because some people genuinely forget and appreciate the reminder. I'm sure you knew all this better than me, but just want to elaborate for anyone who genuinely don't understands who falls for this.
Tepix 2y ago
Your list isn't rendered as a list
kashyapc 2y ago
> Never have an AI agent call someone unless the customer specifically requests that [...] Are there real humans / customers who ask, "yes, have an AI call me?"
- drewcoo 2y ago
  > Are there real humans / customers who ask, "yes, have an AI call me?" If it could go faster than a slow-talking human who needs to re-verify information that someone else already verified? Yes, I'd opt for the AI.
  - kashyapc 2y ago
    True, it depends on the quality of the human on the other side. If it's a mindless front-line support person parroting a script, "did you try restarting the computer?", then I'd take a smarter AI too.
- SteveVeilStream 2y ago
  There are times when I would ask for this. If it's midnight on a Saturday and the website says: "You can reach a human agent 9-5 M-F or we can have an AI agent call you right now" then I might opt for the AI agent. I may also do this if I want to take a call while out for a bike ride, in a noisy place, etc. Talking to AI means I don't need to worry about being polite which is actually very convenient.
marcus0x62 2y ago
> The counter approach is: - Make complete information available as transparently as possible and don't gate it. - Be forthcoming about weaknesses. Don't force prospects and customers to find them. - Ensure that when a prospect or customer does want to talk to someone, they immediately reach someone who can handle the problem or answer the question (no need for escalations.) - Never have an AI agent call someone unless the customer specifically requests that and be sure that all AI agents immediately disclose that they are AI. - Offer flexible e-mail list subscription options (monthly, quarterly, annually, only release notes, etc.) - If the product is not a fit, try to offer something useful anyway such as a suggestion of another product that might be a better match for their needs. From the perspective of designing a sales org to do this it is, roughly, a sales-engineer focused sales organization with the sales engineers on a minimal commission plan (i.e., a pooled 80% base/20% commission plan instead of something like an individual 70/30 plan.) I never saw it in practice, but it would probably have something like a 2:1 ratio of sales engineers to account managers. On the plus side, sales engineers are less costly to employ than account managers. On the down side, this will almost certainly result in lower revenue. > Value based pricing is one of the reasons why a lot of companies end up in these situations. Rather than setting a standard price, the company does a detailed investigation of the customer to try to find out how much value they will gain from using the product and then they set the price based on that determination. Although it maximizes revenue in theory, it is slow and invasive. In my experience (working in tech sales for a few large manufacturers,) this is almost never a routine practice of the sales team. It is sometimes discussed in sales training, and very, very occasionally put into practice, but the Achilles heel is your last comment: it is slow and invasive. Sales people are a diverse group, but they are united by the belief that time kills all deals. Product teams do large scale market segmentation and price sensitivity studies -- that's where you get all of the packaging options. Sales teams? If they have a large existing customer, that customer will already have a standard discount level. That's the starting point for any given negotiation. For new customers or smaller customers, the sales team will try to sell at the 'standard' price, where by 'standard', I mean the prevailing discount level in their region. Variables that will ultimately cause the price paid to change: fiscal year-end/quarter-end, competition, bundling, "incentives" / customer "budget." I put those roughly in terms of how powerful a lever they are to move the price, on average. Fiscal year-end/quarter-end. One challenge with trying to root out traditional sales/purchasing behavior is that customers are trained to reinforce the behavior. While it doesn't always work (a sales rep might be significantly under or over goal and might want a deal to slip into the next quarter/year,) if a customer can force a deal to get done around the end of the vendor's fiscal periods, they have significant leverage to negotiate a lower price if the vendor believes the deal might slip. Competition - this one is pretty obvious, but adding a competitor to your purchase evaluation and letting the vendor you want to buy from know that you are seriously considering their competitor will likely result in a lower price. Bonus points for the purchasing team telling the vendor their engineers like the competitor's product better, but the purchasing team really wants to buy from you because they love you so much. Double bonus points for telling both vendors that. Bundling -- if a manufacturer has multiple products, they often will voluntarily increase their discount level to induce the customer to buy a product 'suite'. The sales reps are often under immense pressure to sell the newer and flakier products on their price lists, and this is usually the only way it gets done. On the plus side, the customer gets more product for the same price. On the down side, the extra products included often aren't worth using, and come renewal time, the true total cost of ownership becomes apparent. "Incentives" / "customer budget" - these are both, typically, BS in that the vendor doesn't usually have a magic "incentive" to offer a one-time discount. The sales rep just got (or thinks they can get) approval for a lower price. Similarly, while customers most certainly do have budgets, they often lie and say they can't buy x because it exceeds their budget, when their internal sales people (the purchasing team) just wants a lower price. I refer to the purchasing team as "internal sales people" because they often-times have a commission incentive just like the vendor sales rep: it isn't uncommon for them to have a personal, individual bonus tied to driving vendor contracts down in price. In some cases, it is a direct commission -- i.e., the purchasing agent (or sometimes budget owner) gets 10% of any monies saved. I wish you the best of luck, but other than taking people completely out of the equation, I think it is really hard to eliminate this behavior in practice because 1) it is so thoroughly ingrained amongst both vendors and customers and 2) it works.
moffkalast 2y ago
The counter approach is better for everyone, less work for the company and a nicer experience for the user, but only in a vacuum. Unregulated competition is a race to the bottom, so the most unethical, invasive, borderline illegal, and manipulative tactics come out on top. Not the best product, but the best marketing. Not the most customer friendly, but the most customer-converting through coercion. Not pricing products and services what they're worth, but the most anyone will be wiling to pay for them.

portaouflop 2y ago

More power to you but in my experience - people don’t read your documentation - if they find one edge-case that is not covered by your product they concentrate all their energy on it even though the competitor product doesn’t even work for the happy path (but ofc they won’t learn that until too late) - you will have a person on the free tier who is not able to the simplest programming take up all the time of your technical support if you let them The rest (disclose usage of AI, offer multiple newsletters (imo unnecessary work), be upfront if it’s not a good fit) is very sensible and works in my experience. The first three will break your product - but maybe I’m just cynical and was holding it wrong all this time edit: I’m assuming you want to build some kind of investor fuelled unicorn - your approach probably will work for a single person slow but steady growth while you have other side income

efitz 2y ago
You need docs. Most users won’t read them. Do a good job anyway because they can save you a TON of support costs if they’re good, even for the few people who do read them. The cost of writing docs is easily offset by even a modest reduction in human support cost, unless you just half-ass your support with LLMs and FAQs and chat with overseas wage slaves. ALSO: your UX needs to warn people about sharp edges and one-way doors, BEFORE your user commits to an action they might regret. These are not incompatible.
- johnnyanmac 2y ago
  If nothing else, the "smart cows" will read the docs and may even help out on support forums when some of the user base doesn't and decides to complain. Docs are good for everyone and will save time internally and externally. >your UX needs to warn people about sharp edges and one-way doors, BEFORE your user commits to an action they might regret. These are not incompatible. Assuming trump doesn't nix it, the recent FTC ruling should fix this by itself. it should be as easy to unsubscribe as it is to subscribe.
  - eigilsagafos 2y ago
    I don’t think easy unsubscribe falls under «sharp edges and one-way doors». But high friction unsubscribe probably goes hand in hand with sharp edges and one-way doors at those lead you to want to unsubscribe…therefore they make it hard.
- freefaler 2y ago
  They also are great for the support team, to "educate" where they can find additional info and for linking as help links if the interface needs this.
- babyshake 2y ago
  Also, it's pretty easy at this point to setup an AI support agent that can reference your docs, and it is a helpful exercise for organizing your own thoughts and being able to communicate them back to yourself. There's not really a good reason to not have good docs. I just assume that if a product is gated behind "let's hope on a call!" it's still vaporware and there's a 50/50 chance that they'll pivot within a few months.

rkagerer 2y ago

Good advice. I created and sold software (and services) for over a decade, and could never have lived with myself if I'd treated customers the way the original article describes.

johnnyanmac 2y ago
sadly, they do it because it works, and (not trying to be too offensive): lotta people are dumb. If you're confused, a nice "team mate" will be more comforting than a cold hard doc and web form. Even if they are upcharging you on a bunch of stuff you will never use. They call multiple times because some people genuinely forget and appreciate the reminder. I'm sure you knew all this better than me, but just want to elaborate for anyone who genuinely don't understands who falls for this.

Tepix 2y ago

Your list isn't rendered as a list

kashyapc 2y ago

> Never have an AI agent call someone unless the customer specifically requests that [...] Are there real humans / customers who ask, "yes, have an AI call me?"

drewcoo 2y ago
> Are there real humans / customers who ask, "yes, have an AI call me?" If it could go faster than a slow-talking human who needs to re-verify information that someone else already verified? Yes, I'd opt for the AI.
- kashyapc 2y ago
  True, it depends on the quality of the human on the other side. If it's a mindless front-line support person parroting a script, "did you try restarting the computer?", then I'd take a smarter AI too.
SteveVeilStream 2y ago
There are times when I would ask for this. If it's midnight on a Saturday and the website says: "You can reach a human agent 9-5 M-F or we can have an AI agent call you right now" then I might opt for the AI agent. I may also do this if I want to take a call while out for a bike ride, in a noisy place, etc. Talking to AI means I don't need to worry about being polite which is actually very convenient.

marcus0x62 2y ago

> The counter approach is: - Make complete information available as transparently as possible and don't gate it. - Be forthcoming about weaknesses. Don't force prospects and customers to find them. - Ensure that when a prospect or customer does want to talk to someone, they immediately reach someone who can handle the problem or answer the question (no need for escalations.) - Never have an AI agent call someone unless the customer specifically requests that and be sure that all AI agents immediately disclose that they are AI. - Offer flexible e-mail list subscription options (monthly, quarterly, annually, only release notes, etc.) - If the product is not a fit, try to offer something useful anyway such as a suggestion of another product that might be a better match for their needs. From the perspective of designing a sales org to do this it is, roughly, a sales-engineer focused sales organization with the sales engineers on a minimal commission plan (i.e., a pooled 80% base/20% commission plan instead of something like an individual 70/30 plan.) I never saw it in practice, but it would probably have something like a 2:1 ratio of sales engineers to account managers. On the plus side, sales engineers are less costly to employ than account managers. On the down side, this will almost certainly result in lower revenue. > Value based pricing is one of the reasons why a lot of companies end up in these situations. Rather than setting a standard price, the company does a detailed investigation of the customer to try to find out how much value they will gain from using the product and then they set the price based on that determination. Although it maximizes revenue in theory, it is slow and invasive. In my experience (working in tech sales for a few large manufacturers,) this is almost never a routine practice of the sales team. It is sometimes discussed in sales training, and very, very occasionally put into practice, but the Achilles heel is your last comment: it is slow and invasive. Sales people are a diverse group, but they are united by the belief that time kills all deals. Product teams do large scale market segmentation and price sensitivity studies -- that's where you get all of the packaging options. Sales teams? If they have a large existing customer, that customer will already have a standard discount level. That's the starting point for any given negotiation. For new customers or smaller customers, the sales team will try to sell at the 'standard' price, where by 'standard', I mean the prevailing discount level in their region. Variables that will ultimately cause the price paid to change: fiscal year-end/quarter-end, competition, bundling, "incentives" / customer "budget." I put those roughly in terms of how powerful a lever they are to move the price, on average. Fiscal year-end/quarter-end. One challenge with trying to root out traditional sales/purchasing behavior is that customers are trained to reinforce the behavior. While it doesn't always work (a sales rep might be significantly under or over goal and might want a deal to slip into the next quarter/year,) if a customer can force a deal to get done around the end of the vendor's fiscal periods, they have significant leverage to negotiate a lower price if the vendor believes the deal might slip. Competition - this one is pretty obvious, but adding a competitor to your purchase evaluation and letting the vendor you want to buy from know that you are seriously considering their competitor will likely result in a lower price. Bonus points for the purchasing team telling the vendor their engineers like the competitor's product better, but the purchasing team really wants to buy from you because they love you so much. Double bonus points for telling both vendors that. Bundling -- if a manufacturer has multiple products, they often will voluntarily increase their discount level to induce the customer to buy a product 'suite'. The sales reps are often under immense pressure to sell the newer and flakier products on their price lists, and this is usually the only way it gets done. On the plus side, the customer gets more product for the same price. On the down side, the extra products included often aren't worth using, and come renewal time, the true total cost of ownership becomes apparent. "Incentives" / "customer budget" - these are both, typically, BS in that the vendor doesn't usually have a magic "incentive" to offer a one-time discount. The sales rep just got (or thinks they can get) approval for a lower price. Similarly, while customers most certainly do have budgets, they often lie and say they can't buy x because it exceeds their budget, when their internal sales people (the purchasing team) just wants a lower price. I refer to the purchasing team as "internal sales people" because they often-times have a commission incentive just like the vendor sales rep: it isn't uncommon for them to have a personal, individual bonus tied to driving vendor contracts down in price. In some cases, it is a direct commission -- i.e., the purchasing agent (or sometimes budget owner) gets 10% of any monies saved. I wish you the best of luck, but other than taking people completely out of the equation, I think it is really hard to eliminate this behavior in practice because 1) it is so thoroughly ingrained amongst both vendors and customers and 2) it works.

moffkalast 2y ago

The counter approach is better for everyone, less work for the company and a nicer experience for the user, but only in a vacuum. Unregulated competition is a race to the bottom, so the most unethical, invasive, borderline illegal, and manipulative tactics come out on top. Not the best product, but the best marketing. Not the most customer friendly, but the most customer-converting through coercion. Not pricing products and services what they're worth, but the most anyone will be wiling to pay for them.