Freelance Bot Building for Messenger Inquiry

(fredrick) #1

Anyone doing any freelance bot building, specifically for Facebook Messenger? I’ve got a few businesses that are interested in having me build them a bot. How are you charging for your services? Thanks in advance for your input.

1 Like
(Samson Taylor) #2

2nd this. I’d be interested to know if anyone has some material on the bot building industry. I’ve been searching for “bots as a service” if you will or bot building but I haven’t found much material on the topics.

Any resources would be much helpful :).

1 Like
(Khashayar) #3

Depends on who the client is and how much work would it take. Our agency charged a hotel 5 digits because we implemented a lot of feature and a lot of NLP. For button-bots, without text-input, it’s under 3K, if there’s no special integrations. I’ve seen 400 dollar internal FAQ bots and 100K dollar bots, for huge brands - so there’s a massive range and it’s highly case-by-case.

2 Likes
(Samson Taylor) #4

@kcfl is there any literature or “industry” standard on how to gauge what we should be charging? I’d love to find a…“starting point” when it comes to what ppl are offering and at what price. This is so new but I think this is a HUGE opportunity to make a lot of $$$ getting in early on the bot building stage.

Your thoughts guys?

1 Like
(Khashayar) #5

I definitely agree. And it’s interesting because in September 2016 we wanted to start doing what we’re doing now and no-one was interested. In less than a year, there’s been an explosion in interest and it hasn’t even gone fully mainstream yet.

In terms of literature, there’s an ebook called The Bot Business that you can read. It’s not a well written book but it has insights on the sales pipeline and those details of running a bot agency. You can get it on Amazon.

I personally think the industry is too new for any kind of serious literature. The first few good quality conversational UX books came out just in the last 5-6 months. It’ll be some time till serious readings can be put together. For now, I suggest keeping up with the existing Medium publications and Facebook groups. Specially the Facebook groups can spark good conversations, be it about methodology or pricing.

2 Likes
(fredrick) #6

@kcfl. Thanks for the insight. Are you doing bot building full time? And, are you doing it yourself or do you have a team? I’m curious because it may be worthwhile to have an offline discussion about possibly working together on some projects (if that’s something you would be interested in).

1 Like
(Khashayar) #7

Yeah for sure! I have a small remote team but mostly for webview and back-end work. All of the bot design and building is me. Would be interested too! I’ll send you a PM!

2 Likes
#8

Nice to hear that someone is building a business around chatbots. We’re trying to start a bot developing business as well, but we’re trying to understand if you can build bots on platforms like FlowXo etc and still sell them to customers this way.

@kcfl the prices that you mentioned are for chatbots built from the ground up or by using platforms like FlowXo?

1 Like
(Samson Taylor) #9

@palmontsagency Using Flow XO and Api.Ai is building from the ground up. It can get very complex when adding HTTP Requests/API’s and the already 100+ software you can connect to already using Flow XO.

Do it your self no tech customers will NEVER touch the depths of Flow Xo on their own, especially if they run a business and have no time to learn all these platforms and software to implement and make work together.

Something that you’ll see most regular DIY’s use will be services like www.manychat.com

Yes, there is the complete start from scratch making a bot from you own lines of code, but IMO that would be a waste of time to not use Flow XO and Api.ai to launch yourself further into production faster.

What are you alls thoughts?

(fredrick) #10

Just wondering…did you get my PM?

(Khashayar) #11

I agree with Samson that this is basically from the ground up. If you had a fairly capable developer in your team, you could have them build a code version of Flowxo ‘actions’ and just use them as modules of a bot. It’d take maybe 50 hours to prepare the environment but the actual building of the bot would take the same amount of time as Flowxo. This means you’re essentially doing the same thing that a developer does when building a bot, with the exception that you have a GUI for it.

Why we use Flowxo over our own coded modules is:

  • Anyone can jump on the platform and make quick tweaks and iterations because of the robust UI.
  • Complexity is easier to manage when you’re not seeing code. Things like filtering and loops, if done on a large scale, can make your code impenetrably confusing. So being able to manage those visually makes it extremely fast and efficient.
  • Some out of the box functionalities of Flowxo that you can implement with 3 clicks, take an hour or two of coding and debugging. Why take a week building the same bot you could build in one day?
  • Reduced debugging time. When you’re using code, a thousand more things could go wrong. So using flowxo reduces your error rate and again, speeds up the development a lot.
  • The SaaS model is perfect for an agency. It sounds weird but since Flowxo forces us into a hosting subscription and takes our bot hostage (no hard feelings), it sort of forces us to do the same to our own clients. We give them a monthly maintenance, updates, and hosting bill and if they want to self-host, we charge them a lump sum that we use to copy their flowxo bot in their preferred programming language. An example would be 35 dollars for hosting per month and 4’000 dollars termination fee, which would pay for a fresh build. Super high volume clients might choose that because 1’000’000 interactions on Flowxo is like 5 dollars on AWS Lambda and it would totally make sense to pay the termination fee and save every month. It’d be really cool if Flowxo had an “export bot” option that you could buy for like 2-5 thousand dollars based on the size of the bot and get the full code, ready to be hosted somewhere else. I think that would be the most explosive feature :+1:
4 Likes