Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: How did you establish and maintain relationships with your first users?
134 points by dwrodri on July 27, 2021 | hide | past | favorite | 61 comments
I'm currently working on a side project and will be searching for early users / alpha testers in the near future. Our target demographic is established, but we want to keep an open mind as we iterate on the product. Anyone have any "dos" or "don't"s that you can share? Also, sharing a brief description of the product/service itself would be really helpful as well.

For context, I am working in the computer vision space.




I'm going to give a fun alternative example.

I used to run some popular publishing websites around entertainment: gaming, movies, tech...etc.

I talked to our first users entirely through our message boards, which had PM systems built in. The regulars were there almost every day and I got to know them fairly well. We'd talk daily and I used the message board to show off new designs I was working on. Eventually they started their own fan podcast (this was 2008 or so mind you) and I'd show up as a guest to support their love of the websites. They'd be surprised I'd show up. Me? I was surprised they went to so much trouble to support my site.

Eventually we just hired a bunch of these early users at interns. One we hired permanently, and others we helped get jobs within the gaming industry. I don't see them that often in person anymore (I've since moved) but I'm happy to see them with their own thriving careers and consider a couple of them friends.

I sometimes wonder if the move away from message boards and towards larger communities like Reddit and Twitter would have allowed for this kind of clubhouse atmosphere. It's one of the bits of the old Internet I occasionally miss.


> I sometimes wonder if the move away from message boards and towards larger communities like Reddit and Twitter would have allowed for this kind of clubhouse atmosphere.

I occasionally use Reddit as my forum / message board for whatever project I'm working on. Just set up your own sub. Works pretty well. Reddit can be as small of a community as you want it to be, if you control the sub. Granted, you're operating within Reddit's control, so something could go wrong there. However along with such tradeoffs there are aspects in your favor as well (such as the tens of millions of active Reddit users, far lower friction adoption of your community board, not having to manage forum software, and so on).


For an example of this, see r/supernote.


I saw the username and assumed giantbomb, big fan :)


I'm the founder of https://www.scrapingbee.com/, an API for web scraping.

At the beginning we had a banner on the dashboard with a CTA to book a call with me. To incentivize users to book the call, we offered free API credits: https://twitter.com/PierreDeWulf/status/1329765250707623939/...

I had 100+ calls with users in the first months, it helped us a lot to understand our market, what they needed etc.

It worked so well that we had to remove the free credit incentive because taking those calls took all my time.


But I would guess that a significant number of your users are doing illegal web scraping (Otherwise, why use proxy rotation). So you advise them how to get around website protections?


I'm not a ScrapingBee customer (but I've looked into and was interested in it). A lot of websites have started blocking any datacenter IP, even if you are just scraping it once. In order to get the contents of those sites, you need to use residential proxies.


When is it illegal to scrape a website?


There are bad scrapers out there. Plenty of common problems:

- Denial of Service by queries - they hit search pages with complex or slow queries, diving to 1000th page of results . This kills the db.

- Denial of Service by parallelization - they hit 1000s of pages at once, causing server to run out of memory or other issues . This kills the web.

- Denial of Service by bugs - their code is buggy, slight change to page causes their scrapes to repeat ad nauseum.

- bad URL/cookie scrapes - they hit URLs that perform actions (say add to cart) against websites. This causes sites to track more data in abandoned carts, managing sessions, item popularity.

If scraping wouldn't affect server negatively, then it would feel less illegal.


Let's not forget data mining. People build whole businesses on this and a lot of them are parastic. They are profiting on data that is sometimes costly to obtain.

All the companies that scrape linkedin tied it to other socialmedia and build power profiles on people such that CIA is clapping with smile.


How is it relevant to the topic of "establishing and maintaining relationships with your users" what the users are doing by using the service in their jurisdiction?


It can be relevant, though I don’t know whether GP meant this.

It’s usually much easier to get users for shadier services, because they have a dearth of good options (since not many legitimate businesses are interested in addressing the problem).

So much so that the lessons learned in this space may not be applicable at all to the market as a whole.


I'm not sure if all web scraping that's undesired by the website you're scraping is illegal.


Been working on Newsy for some time.

https://www.newsy.co

Newsy is a tool that converts your un-used domain names into a Reddit-like content aggregator. Allows you to make your domains useful rather than just sitting there.

We did an alpha launch through Betalist and had ~250 people sign-up. From there on, we emailed and talked to every single person who would reply to our message. That built pretty solid relationships which still last today.

Something surprising I realised was that people engaged more when I asked them about who they were (e.g. What do you do? What do you do for full time job?) and when they replied, I kept asking about their role, challenges. Sometimes I ask about where they are from, if they are ok from COVID. I was genuinely curious about those users as humans (how did you find my tool? What do you do on the other side of the world from me? Is it cold? Hot? etc etc).


Dos:

1. Buy a tollfree phone number and list it your website. Some people like to talk to a support agent rather than email.

2. An official email makes the business appear legitimate. Have an email@companyname.com.

Don'ts:

1. I didn't have good experience with website chat widgets as I found them confusing to configure and that generated numerous low quality interactions.

2. Since the volume of emails is low initially, better to avoid email automation as that can take up a chunk of time to set up. Start with manual emails and once you know which email copy evokes a response, you can think about automating them.


We (<20p SaaS) use Intercom with great success. It's hooked up to our support email as well, so users who prefer email can use that, while giving us a central mailbox for it. Most customers really appreciate quick, low barrier chat. We all take turns as operator (we call it Empathy Day).

As you handle support for a while, certain recurring topics appear. We identify these and either use canned responses (macros), or (preferably) improve our product to prevent these questions in the first place. We also have a bot for some basic things like pointing at the docs for certain topics.

We sometimes are asked to jump on a call, but only do that for highly technical support requests and enterprise customers. Since we have customers worldwide and a globally distributed team, synchronous communication is tricky anyway.


How did you decide on Intercom over the dozens of competitors in the space?


This makes a lot of sense: Ideally, we'd want a small (<20) amount of "future power users" as our focus group, since we imagine these users would provide us with the most most powerful growth via word-of-mouth if we work extra hard to make that initial focus group happy.


What worked for us:

We did something like this with our early users. Called the thing "The Builders Program"

For the medium, have one default you use but don't limit things to it. If one of your early users wants to use twitter to contact you and give you feedback, use that. We created a Discord server, had a group Twitter DM, used Email.

One thing you should do early on with these users is show them you are listening. Pick something they say and ship it as soon as you can. If they see you're listening, they'll be more active in sharing feedback.

In exchange for being part of the builders program we gave them lifetime free access to the product later on.


Yes, I think that valuing them through a named program is a great idea.

... and finding out what 'reward' they value the most, it might be different for different folk but free access to the product they supported is a great start.


Hi, I am Victor founder of https://webautomation.io/, our tool helps companies extract data from the web using pre-built webscrapers

Here are my top do's and donts solely based from my experience Do's 1. Experiment with paid ads,its probably cheaper than creating organic content 2. Have a chat bot, and use this to have open conversations with visitors and users 3. Find any excuse to get people on a call i e demo,customer support etc 4. Yes your target demographic might be clear to you, but still listen to the data, I thought I was building for data engineers but alot of my users are surprisingly marketers

Don't 1. Copy other people's strategies, it doesn't always work, instead experiment different ideas and see whats working 2. Outsource any early interaction with customers, its best you do yourself and get a feel for what people want

Good luck please share updates and let us know how much progress you made


> Outsource any early interaction with customers, its best you do yourself and get a feel for what people want

Would you mind clarifying that statement, or giving some more details?

Huh? That seems like a contradiction. "Outsource" implies that you aren't going to do it yourself.


That one was in the Don't section.


It’s on the “Don’t” list.


A community Discord has worked well while building Print Nanny, which monitors 3D printers using computer vision. https://www.print-nanny.com/

I love that people share their latest projects, chat about experiments. 3D printing is such a creative hobby and I'm constantly amazed with the ingenuity I see. Really helps me stay motivated through the night/weekend grind.

A few stats about my beta program:

* Roughly 2,200 applications for the private Beta program

* 200 beta invitations sent

* 100 actually used the product

* 180 hanging out in Discord

* 6 months of nights/weekends development

* 10 paid Beta spots offered, sold out in 2 days

I do support over email or Discord.

I think event-based analytics like Posthog are more valuable than a chat bubble integration, if you have to choose 1 to implement. People often tell you more with their actions than with words.

(Edited list formatting)


I'm building a public discussion site (https://sqwok.im). So far I've learned a few things:

1) Most people will happily give you feedback. Some of it will be good, some of it you won't agree with, but keep an open mind, and always thank them.

2) I go out of my way to tell new users that I appreciate their taking the time to check out the site.

3) If you get the opportunity, engage with your early users and talk on a human level. Many times that simple act of connecting and having a conversation with another person will spark an increased interest in the success of the project and you'll find that people will reach out to you when _they_ have an idea or suggestion.

4) List your email and/or a phone number if applicable, and likewise if you're having a good conversation with someone, ask if they'd be ok with you emailing them with some follow up questions. I've had a few people whom I got to know a fair bit and have exchanged multiple emails.

5) Lastly, I've made one specific mistake that I'm learning from, which is to _not_ neglect your early users. That is, when people start to show interest in your product, make sure to show up each and every day to interact with them and keep the interest up. I've had a few periods where I went into a development k-hole for weeks on end and wasn't as active on the site, which led to a decrease in overall activity.

Good Luck!


> a development k-hole

Two terms I didn't think I'd ever see used together


haha, well, that's how I can best describe when I get hyper-focused on something and tune everything else out... it can be unhealthy sometimes and throw my sleep schedule off.

clicked your prof, nice site! where do you surf in SF?

https://brian.team/projects/laser-cut-audio/ very cool!


We're an early stage startup [0] for web analytics and something that worked incredibly well so far is Discord. A lot of our customers/users have a Discord account and it's very easy to join a server and get into a (high quality) call, or to use the chat. That's something I didn't expect to work so well.

Appart from that, I like to ask questions in email replies and give customers the feeling to have influence on our development process (which, in fact, they have). Even if someone doesn't become a customer or signs up for a free-trial, you can learn something from them.

[0] https://pirsch.io/


Little OT, but absolutely love how your landing page looks and feels. I’ve seen a few similar ones - is this built from a template? If so, could you please share it?


It's all custom build :)

[Edit] Adam created a clone of it using Tailwind: https://www.youtube.com/watch?v=R7dob2GGt70


Awesome! Thanks for sharing


There are way too many CTAs and general elements on the top of the landing page.


Yeah, we're working on an update reducing the content a bit.


If your users sincerely have a pain point and think you can solve it, they will usually be happy to give you lots of input and feedback. This requires two things- intending to solve a real problem that users care about and being credible. You can increase the second aspect by communicating with and listening to your core and ideal customers, and getting to know not only if the problem matters, but what they care about in the context of that problem. There's a maxim that the value of your product is what your customers think it's worth. Learning from them why they care will increase your chances of building strong early relationships.

Mechanically, regular communication that offers value via email, plus actual conversations that are informational not sales-y.

From being the person doing this at both B2B and B2C companies in early stage.


Shared Slack channels.

We're an early stage e-commerce startup selling mainly to medium sizes companies, and one thing that's worked incredibly well is setting up a shared Slack channel with each customer's org. They can get hold of us quickly, we can both pull relevant people into any conversation, and it makes the relationship feel so much more collaborative because it's almost like you're a part of their team.

We've gotten a lot of positive feedback about this, and being available to fix any issues is a great way to get over some rough edges in the platform you are bound to have early on.


Assuming your users are also users of Slack, I highly recommend creating shared channels; that way, they can reach out to you in your slack, without leaving their own. Email is great, but there is something about IM that feels way better.

Also, if you are offering your product/service for free initially, I would say, "it's free, but in exchange you gotta give us 20 minutes on a call every week/every other week". This ends up being super valuable.


Depending on the target demographic, Discord might also be a good Slack alternative. It seems like its the preferred platform for many projects nowadays


As having been on both sides of the relationship building effort, my vote would be for just email: quick, scrappy and personal.

For example Taylor from Focusmate once reached out to core users by just saying „hey, it‘s me, Taylor, not a marketing person, I‘m building this thing and you seem to be using it a lot, do you have any feedback?“

Interestingly enough, this approach was as rare as effective: I just gave him some and he thanked me.

I just wondered why not more people did simple stuff like this.


I just got a phone call yesterday from Autodesk, saying they’ve seen me active using the new Generative Design function, and asked how I’m finding it / what I’m doing with it to make money. Great feedback for them, and I felt chuffed to get a phone call, and am happier with my software than I was before - win win!


Glad you liked that... but they knew which features you're using via telemetry?

I would have probably given them an earful, including a demand for a refund.


I did a public chat. It was a pain to code and a bigger pain to moderate. But it worked extremely well. A side effect is we never got bad ratings because anyone could just make an anonymous account and rant and complain where we'd see it. We ended up with two WhatsApp groups that lasted longer than the app itself.

I tried doing a Discord server for it recently but it doesn't have the anonymity effect.


your product was a public chat? how did you market it/what was your plan for it?


No, I made a chat room to communicate with users.

The product was a diet recipe app. The users decided to create WhatsApp groups to support each other in sticking to a diet. I had no trouble marketing the app though. I just posted a link on Facebook and everyone shared it. We got 1,200 downloads in the first 24 hours.


> We got 1,200 downloads in the first 24 hours.

that's impressive!


Reaching out to them manually on Twitter. DMs are surprisingly effective.


I'm working on https://www.monsterwriter.app/

Every know and then I get an email with feedback or feature requests. Those are good opportunities to start a chat or a discussion with the user. Before I do that, I try to build what they want. Once Shipped, I write them a follow up email.

The other thing I tried is to reach out to people who I might think are interested in my product. I explained them that feedback is the most crucial think for me at the moment. This worked out quite ok. Some people took the time and had a look at the app and actually answered my question. I have to say that I only reached out to people who have the same mother tongue as me. In this case we already have something in common and it is more likely to receive an answer.


I'm working full time on a project substantially less technical than computer vision (https://www.reportcardwriter.com) where I'm still getting early users.

I haven't found any method more reliable than tapping my personal network and conducting user interviews with actual human beings who teach elementary school.

I'm still in the very early stages of setting up content marketing (https://blog.reportcardwriter.com), and I'm still iterating on the final design of the software, so I haven't really seen any inbound leads yet. I have a longer term strategy for that, but for the moment, I've gotta hunt for everything I kill.


We usually run a survey with users who join our telegram community with one question (2 at the max) only after landing significant feature updates to the service.

We also send (often a different) survey email (along with product updates) to folks who've subscribed in case we expect to learn something else than from the responses in the telegram group. In our experience, with our audience, emails don't work as well as Telegram.

What we also do to gauge user-interest and build a community for an Android (userspace) DoH client and firewall we build is:

1a. Include a link to Telegram community in the about page prominently.

1b. After every update, shows a changelog to the user with a button "suggest features" that emails feedback.

2. Contact email features on all our webpages.

3. Google the past 24 hours for keywords related to our app to see who's discussing it, and if it is a forum like reddit, leave a comment and personal email (as opposed to generic support at domain dot com) encouraging the user to contact.

4. Newsletter to collect potential customer email-ids. As mentioned above, survey are always sent out to them once in a blue-moon and only with accompanying significant product updates.

5. All incoming to generic mailbox are replied-to from a personal email. Emails always end with a remark asking the user to not shy from sending bug reports, queries, and suggestions.

6. Monitor related subreddits where similar products/apps are discussed, and post replies only after the thread is off the front-page to avoid derailing a hot discussion with self-promotion (which gets you banned on most subreddits).

Provide white-glove service to users and they remember you for it and usually champion your product henceforth, which helps when the user base is young and the product is rough around the edges.

First answered here: https://news.ycombinator.com/item?id=26754095


We use Intercom chat which is connected to our internal Slack at Bugasura.io (we are building the modern bug tracker)

It is quite difficult to get to talk to the users unless they have technical query.

Some things that work for us

1. Having the chat helped moved people further up the funnel 2. The chat offers to help at different points in the product (you can setup rules in intercom to bring up the chatt) 3. We also talk to paid users via email proactively (not many at this time) and funnel their input to the sprint


We built a low-friction discussion forum for our price comparison site (yes, NIH syndrome). It's still used for feedback 21 years later, though e-mails, feedback buttons on the website were added too.

In general, most B2C product feedback probably exposes UI/UX weaknesses rather than desirable features and adding more features instead of making the existing ones work smoother and faster is usually not appropriate.


Mine is an online animation editor that requires no coding or formulas https://toonclip.com

The first users came from a post here on HN and Product hunt.

For community building we have a Forum and a Discord server. Also our animations can be exported and shared on social media. As these animations have the ToonClip log,o it drives a decent bit of traffic.


I just have free apps, so I didn't do much. There's an email they can contact me at. I have recieved a few requests for enhancements through that.

I also run a honey business. I tell people to follow our facebook page so they get an alert when we have a harvest since quantity can be limited. I'm also working on a website with an email list.


My product is a web B2C SaaS that will be initially targeted at hobbyists/beginner/intermediate programmers, so I think getting them to believe in the product will either require a killer demo or a free tier they can play with which provides them with value.

Do people really just start showing up if the service has a free tier? I'd assume you did some basic marketing.


The free apps I have are not very popular. I did try some basic marketing, but it wasn't that successful.

I would guess the best marketing you can do for a developer tool is to create tutorials around whatever problem you tool solves. I know I tend to stumble across stuff when googling "how do I do...".


this is admittedly self-promotional but i've built a free app that helps with this... it's a web app designed to literally "maintain relationships with users" especially new projects.

you can see how i'm doing this in www.yen.chat — the communication platform is also a community platform where i can get feedback about the product itself, from them directly. i engage and build community around topics that we all like and slowly build the project.

i'd love to help!


It may seem simple, but list an email address on the website and invite people to use it. Whether or not you have a DIY free trial or signup, you’ll want to make it easy for interested parties to contact you with questions before starting to use your project.


Get to know them. Some might be wackos but just limit your time, not your conversations.

You want to be open to radical departures from the current implementation. You may think you know everything but before you have customers you actually know nothing


This is going to depend so much on your product and target demographic.

We cold called.


1) How did you get their contact details?

2) Did you have any more success with cold calling vs cold emailing?


1. We were able to figure out a strategy to Google them, then we outsourced it to another country.

2. Cold calling worked better when we had a really high conversion rate, but was more time consuming. Which worked best would depend on the price of the sale, and your conversion %.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: