Massive Scale Cross-Border Amazon Arbitrage

[GoldFibre]

​

Weekly Goals​

• Follow new routine and work at least 20 hours (15 / 20 hours)
• Take care of all the major things on my bot to-do list (5 / 5)
  • DONE Ensure queue is working properly
  • DONE Fix destination shipping calculation on orders
  • DONE Re-calculate cost on recent country 2 orders
  • DONE Handle blacklisting in recommended ASIN queue
  • DONE List 'safe' competitor ASINs

5-7am​

The initial fetch for running my queue was taking 20 seconds, so I found the root cause and now it is instant - this will speed up my bot a great deal.

5-9pm​

I focused on the non-tech part of things:

• I followed up on trying to source products for my +200k USD wholesale inquiry. On Tuesday I had called several offices who all pointed me somewhere else and eventually to email. I emailed several addresses and got no reply. So today I went on LinkedIn and messaged every salesperson at the company I could find with the same message to see if anyone in their company actually wanted to sell something. All in all I messaged 20 different people. If this doesn't work I will start messaging people outside of the sales department.
• I contacted seller support to ask them what documents I can submit, since they rejected my utility bill. There is no instant call feature for this country, so I requested a callback.
• I asked my account manager if there is anything he can do about my BuyBox eligibility and he put in a request - he said we should get a response in 1-2 days
• I followed up with the brand I had wanted to sell but who didn't want me listing on Amazon
• I set up email against my Shopify domain and updated it on the store, because one of the guys I messaged on Fiverr said this could be a reason my GMC applications get rejected. I added a logo to the Shopify store, again because the guy on Fiverr said this could be an issue. Then I messaged the Fiverr guy again to negotiate him down, because he quoted me 300 USD.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Bounce Back]

The amount of effort it takes sometimes to get someone to sell you something is so unreal isn't it?

 

[GoldFibre]

The amount of effort it takes sometimes to get someone to sell you something is so unreal isn't it?

I only have one reply so far, and all he sent me was a thumbs up emoji!

 

[GoldFibre]

Weekly Goals​

• Follow new routine and work at least 20 hours (18 / 20 hours)
• Take care of all the major things on my bot to-do list (5 / 5)
  • DONE Ensure queue is working properly
  • DONE Fix destination shipping calculation on orders
  • DONE Re-calculate cost on recent country 2 orders
  • DONE Handle blacklisting in recommended ASIN queue
  • DONE List 'safe' competitor ASINs

5-7am​

I got a couple of replies on LinkedIn to my wholesale inquiry. Most of them were 'I'll get back to you'. One guy asked me to email him, so I did. I replied to everyone else.

I received pictures from the factory in China that is making the samples. I thought they would have been here already, but I guess they needed to produce the order first.

9-10pm​

Transactional work.

I sold a knife, which I can't import. I have the word 'knife' blocked if it appears in the title or the description, but not the product type 'KNIFE' and so this knife passed through the filters. I fixed that.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Bounce Back]

View attachment 54658

Weekly Goals​

• Follow new routine and work at least 20 hours (15 / 20 hours)
• Take care of all the major things on my bot to-do list (5 / 5)
  • DONE Ensure queue is working properly
  • DONE Fix destination shipping calculation on orders
  • DONE Re-calculate cost on recent country 2 orders
  • DONE Handle blacklisting in recommended ASIN queue
  • DONE List 'safe' competitor ASINs

5-7am​

I got a couple of replies on LinkedIn to my wholesale inquiry. Most of them were 'I'll get back to you'. One guy asked me to email him, so I did. I replied to everyone else.

I received pictures from the factory in China that is making the samples. I thought they would have been here already, but I guess they needed to produce the order first.

9-10pm​

Transactional work.

I sold a knife, which I can't import. I have the word 'knife' blocked if it appears in the title or the description, but not the product type 'KNIFE' and so this knife passed through the filters. I fixed that.

How do you find out they knife got through and isn't okay on import rules/laws? Curious on how you prevent yourself from accidently breaking these restrictions with so much automated.

 

[GoldFibre]

How do you find out they knife got through and isn't okay on import rules/laws? Curious on how you prevent yourself from accidently breaking these restrictions with so much automated.

I got an order that said 'Knife' in my fulfillment spreadsheet.

I probably prevent 99.9% of them, but this one was an edge case (pun intended).

 

[Bounce Back]

I got an order that said 'Knife' in my fulfillment spreadsheet.

I probably prevent 99.9% of them, but this one was an edge case (pun intended).

So if someone reviewing that sheet didn't know the laws it would slip through? I am trying to think through your future scaling when someone who cares less than you may have this responsibility of review (but you can always solve that later haha).

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

So if someone reviewing that sheet didn't know the laws it would slip through? I am trying to think through your future scaling when someone who cares less than you may have this responsibility of review (but you can always solve that later haha).

If my bot doesn’t catch it then I will list the product. If I don’t catch it during procurement then my vendor will check prior to shipping from the US. If my vendor doesn’t catch it then there is a chance that customs inspects and catches it.

The later in the process I find out, the more cost has been accrued against the order. Also, cancelations count against my account health score, so I need to keep them low to stay in business.

 

[GoldFibre]

​

Weekly Goals​

• Follow new routine and work at least 20 hours (19 / 20 hours)
• Take care of all the major things on my bot to-do list (5 / 5)
  • DONE Ensure queue is working properly
  • DONE Fix destination shipping calculation on orders
  • DONE Re-calculate cost on recent country 2 orders
  • DONE Handle blacklisting in recommended ASIN queue
  • DONE List 'safe' competitor ASINs

9-10pm​

The Shopify app I was using/abusing to import lots of Amazon products into Shopify added a monthly cap and a high usage fee on it, so this option won't work for me anymore.

The past three days I have been sick and was still going non-stop, so I took a rest day and am feeling a lot better.

I only did my transactional work at night.

 

[GoldFibre]

Weekly Update​

I hit my goals for the week early on. I made a bunch of improvements to the bot.

I worked 19 out of my goal of 20 hours for the week, including working several days while sick.

There wasn't much progress on other revenue streams:

• The marketing agency idea with my friend is going in slow motion. This is the issue with partnerships. Each person can have a different idea of the speed/effort they want to put in and that can lead to frustration. For this thing, if it ever goes forward, it will be 100% outsourcing. I definitely won't be coding anything myself, and only will lend my expertise.
• I'm trying to hustle to source product for my 200k+ USD order but still haven't found anyone I can get on the phone and talk to. Even if I find the product, I won't have the money to finance the order, so there is another obstacle to overcome.
• I made some improvements to the Shopify store, which I'm hoping will get me approved by GMC and Google Ads once I can reapply on March 20.
• I have other ideas floating in my head and some bits of code that I've written, but haven't taken anything forward.

This week I want to again spend 20 hours working, but primarily focus on getting some potentially revenue-generating idea up and running by the end of the week.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

Weekly Goals

• Work a total of 20 hours (4 / 20)
• Get a new revenue stream up and running

5-7am, 1-2pm, 9-10pm​

I spent the morning brainstorming ideas and looking at resources around building a SaaS. For sure I can code everything myself, but that would take a long time. To reduce the time, I am particularly interested in boilerplate I can use for user subscription management and other things that every SaaS has. I found one project that appears to have a lot of what I need and using the same stack I'm most familiar with - React/Next/MySQL.

I wrote down a number of ideas of things to build and rated them by level of effort. I plan to start with something simple in terms of functionality so that I can get it out of the door and learn as quickly as possible.

 

[GoldFibre]

Weekly Goals​

• Work a total of 20 hours (7 / 20)
• Get a new revenue stream up and running

5-7am, 9-10pm​

I normally drink the equivalent of about 5 espressos every day. Yesterday I quit drinking coffee cold-turkey and had a monster headache and felt sleepy the entire day. Today I was much better although still a little groggy. I think I'll be fully through the withdrawal symptoms by tomorrow.

I 'worked' by watching some YouTube videos about building a SaaS, although I was probably indulging a bit in being lazy. I also let some of the videos get to me and discourage me a bit.

Later in the day I decided I will just build what I can within a week and see how it goes. The next week I'll either build something else or continue working on the same one if it seems promising. I figure that if I had worked like that every week starting a few years ago I would already be rich and retired by now. Worst case scenario I end up making a YouTube video next year: 'I built 52 SaaSes in 1 year made no money'. This doesn't even seem like a possible outcome.

 

[StrikingViper69]

SaaS doesn’t need to be complicated - someone here posted a while ago about a web version of Mail merge that was making the developer 5 figures a month.

Are you wanting to move away from your current arbitrage endeavour?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

Are you wanting to move away from your current arbitrage endeavour?

I'll keep doing it, but the growth here is limited because I can't use the ASINs that I mine automatically anymore. It's too risky and my account will likely get permanently shut down.

So I need to find different income sources before I can really escape from my job.

 

[Bounce Back]

I normally drink the equivalent of about 5 espressos every day. Yesterday I quit drinking coffee cold-turkey and had a monster headache and felt sleepy the entire day. Today I was much better although still a little groggy. I think I'll be fully through the withdrawal symptoms by tomorrow.

I personally gave up caffeine a few years back and it was a smart move. A couple times over the last few years I had even half a cup of coffee and instantly remembered why I stopped. It is hard for some weeks but it goes away - I've found starting the morning by chugging a glass of water helps.

Though I wasn't "5 espresso" tier - that is next level kinda stuff!

 

[GoldFibre]

Weekly Goals​

• Work a total of 20 hours (12 / 20)
• Get a new revenue stream up and running

5-7am​

I took a prebuilt SaaS example from GitHub and worked to connect it to the various services that power it, set up the database, etc. My idea is to modify this into what I need because the vast majority of the boilerplate is done: billing and user management

Once this boilerplate is in place I just need to put in the parts that apply to my service.

5-8pm​

In the first two hours I was able to fix the bugs with this repo and get everything running on my Stripe test account. So you can sign up for a subscription and it will create the monthly billing in Stripe.

Then I went and edited the landing page to reflect what my service is supposed to do (get you indexed by Google). Here is what it looks like so far with a quick Canva logo and some text changes:

9-10pm​

Transactional work. Country 1 BuyBox eligibility is back finally, and this is why there are more orders today.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (17 / 20)
• Get a new revenue stream up and running

5-6am​

I worked on the actual functionality of the SaaS, investigating how to get permissions to query the user's Google Search Console data and make requests to the Indexing API on Google Cloud. I have a rough idea in my head now of how this will work and what the interface will look like for the customer. I also have the rate limits now, which will affect the capacity I put in the subscription tiers.

I had slept pretty late the night before, so I slept again at 6am and woke up at 7. I didn't want to be a zombie when I worked on this again in the evening.

5-8pm​

First I got the GSC sites for the user to display in a list.

After that I spent an hour trying to figure out something that turned out to be a bug in Clerk. But I found a workaround online and moved forward.

Then I made a simple form that lets you type in a URL and then it will attempt to submit it for indexing. This worked, so my product actually functions now.

The better user experience will be to display URLs that are not yet indexed, let the customer submit them with a button press, and display their remaining quota for submissions on the screen. I'm not sure if there is an efficient way to get the unindexed URLs. There is definitely an inefficient way of iterating over all the pages in the sitemap, but I hope I can find a better way.

Finally I added success and error toasts after submitting the form.

9-10pm​

Transactional work.

 

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (22 / 20)
• Get a new revenue stream up and running

4-6am​

I worked on the plan-limiting part. So depending on the plan you get a certain number of URL submissions each month. Now it displays the number of submissions remaining on your account for the month and it will return an error if you try to go over that limit when you submit.

Then I put some styling on the form:


This has the quota hard-coded to 10 per month, so next I wanted to connect the quota to the active plan of the user. But it looks like this code has some bugs and the DB is not update properly after submitting the payment to Stripe. I found the issue and fixed it so that the DB is up to date.

Now the limit is linked to the active plan:

6-8pm​

My goal was to get the entire OAuth registration and subscription flow running perfectly.

• I removed the email signup option because you need a Google account for this service
• I fixed the Sign In redirect to go to the right page and skip the user creation form
• I found a bug with signing out and then signing back in - the Clerk sign in dialog doesn't appear. But I'll treat this as an edge case and handle it later
• I got distracted researching how to check if URLs are indexed and found a way to do it
• I made some badges for the current plan and a button to upgrade

• I created an upgrade page and tested it. It didn't work
• I simplified the DB schema to store the plan against the user directly
• I made the upgrade page redirect to the billing page and defaulted the plan that was selected for upgrade

9-10pm​

Transactional work.

 

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (28 / 20)
• Get a new revenue stream up and running

5-7am​

Since I only have two days left, I focused on getting the site up and running in production.

• Connected the repo to Vercel
• Created a serverless PostGres DB and connected it
• Switched my schema from MySQL to PostGres
• Fixed a bunch of lint errors
• Started setting up Stripe for production

5-8pm​

• Set up my Stripe products
• Tried to publish my Google Cloud App and discovered more to put on my To-Do list

• Made a privacy policy by copying the template from Shopify and then removing all the references to Shopify
• Wrote a one page document explaining why I need access to the user data
• Bought the domain indexfast.app on Vercel
• My site still hadn't successfully deployed so I fixed a half dozen things and finally got it running, then assigned the new domain to the deployment
• Added a link to the privacy policy in the homepage navigation
• Verified my domain with GSC
• Since the video needs to be taken on the production application I then submitted for verification

Looks like it might be a while before it gets verified.

But in the meantime my site is live and you can actually log in with Google: www.indexfast.app

• Fixed one issue with the DB being tied to the test values
• Updated Clerk to production
• Made some UI improvements to the form with loading state and decrementing the remaining submissions
• Asked my friend to try it out and he gave me some helpful feedback
• Looked at some competition to benchmark pricing and quality of the site. Here is one: Go index me!

9-10pm​

Transactional work.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (33 / 20)
• Get a new revenue stream up and running DONE

5-7am​

I reviewed some competitor sites that do the same thing.

Go index me!

Makes sure all my pages are indexed by Google

goindex.me

I explored that competitor site a bit. His landing page is really nice, but once you get into the app it turns out there is no free version that is promised, the UI is broken in a few ways and the product is a bit confusing. He has 30k+ backlinks but estimated traffic under 10 visitors per day. It turns out he signed up for some affiliate service and this is where all the backlinks are coming from. I guess this doesn't guarantee much traffic.

I don't know if this is making any money or not

INDEXED.pro | Fast + Auto URL Indexing Tool for Google SEO​

This is a team of one tech guy and one marketeer. They had 50 customers pre-launch. Maybe those came from running ads to a landing page to collect email addresses. They spent two months building it, which seems about right with what I was able to do in a week.

Their X account doesn't have much traction, so I am assuming that most of the customers came from paid marketing.

I think their landing page is even better than the first competitor. However, they don't have any free version or show you anything about the product outside of the landing page. Once you log in they just put the subscription buttons in front and every link points you back to them.

Is this a good or a bad thing? On the one hand you are pushing the issue to get the customer to pay, but on the other hand you don't really know if the service will work or not, so you have to go by the social proof on the homepage.

Both sites rely on a service account key from Google Cloud, which is a pretty techy thing to require. In the FAQs they said it is required to make the service work. It would be great to find a way around this to make it easier for average people to use the product.

They have some ambitious plans to build a bunch of SaaS products this year. They have posted this table:
The table doesn't show that their first product was a flop and didn't make any money. This is assuming that every product afterwards will make money, but not as much as their indexing tool. I hope they can do it - they will definitely be on the fastlane if they can!

My Site: indexfast.app​

When I started looking at these sites I was initially overwhelmed by how good the landing pages were.

But now a few things reassure me:

1. The product is not that sophisticated - I can build it
2. I have a couple of ideas for improving on both of them, mainly the onboarding experience
3. Someone is actually making money doing this and with a similar time investment as mine

So I think for this idea to work I need to drive traffic somehow and make sure the product onboarding experience is so simple a caveman can do it. Maybe if the service account key is needed to make it work properly, there is a basic version that works without it, so the customer gets a feel for it. Then you tell them "In order to make this scale 100x we need to follow these steps." and walk them through it.

Another idea is to do this in different languages. Both of these sites only work in English, which is the biggest market, but maybe there will be less competition with a site in a different language.

I wonder if the first competitor I looked at is an example of a tech guy building something and then giving up or not putting enough marketing behind the idea. The products are essentially the same, but it seems like the second one is doing much better. The difference is really the marketing.

9-11am​

Using the competitor as a benchmark I redid the landing page.

Here is a before and after screenshot:

8-9pm​

Transactional work.

 

[GoldFibre]

Weekly Update​

Weekly Goals​

• Work a total of 20 hours (33 / 20)
• Get a new revenue stream up and running DONE

For the first time in a while I have hit all my weekly goals.

I was debating if I should write a second SaaS in the next week continue on this one, but I think it's clear that this one needs more attention before I can say if it works or not.

I'm going to work on IndexFast with a goal of getting my first paying customer. In order to do that I need to improve the product a bit more and then either sell it to someone I know or try to convert someone with digital ads.

 

[theag]

make requests to the Indexing API on Google Cloud

I'd be careful with that. This API is only intended to be used for job postings and livestream announcements. So your service is most likely violating Google's TOS, which usually isn't a good foundation for a business.

I'm in some SEO groups and heard about recent manual penalties against sites that abuse this, so it seems like Google is starting to crack down.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (4 / 20)
• Get first paying customer for IndexFast

4-7am​

If I want a paying customer, then my payment links need to be working. If you click on them from the home page, then they don't work because the billing page was an auth page.

This led me down a bit of a rabbit hole and I had to fix a lot of things in the flow that didn't really work properly with the template I started with. I didn't fully solve it yet, but I'm much closer now and will have a simpler logic for users and auth and payment.

9-10pm​

Transactional work.

 

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (10 / 20)
• Get first paying customer for IndexFast

5-7am​

After about an hour I got the Stripe payment form rendering again. But I had more issues and refactoring to get through and didn't finish it all.

5-8pm​

After a lot of refactoring and learning about Stripe workflows and webhooks, I got the basic sign up (free plan) -> upgrade flow working.

There are still a bunch of other sub-flows to work through, but after today I feel like I have a much better handle on how everything fits together.

9-10pm​

Transactional work.

 

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (16 / 20)
• Get first paying customer for IndexFast

5-7am​

Since the 'happy path' flow is working and my goal is to get my first customer. I think I have enough for that. So now I'll work on other things I need to get that first customer:

• Fixed the sidebar nav to show the pages available
• Improved the initial screen by adding 'Unverified' badges when someone has a GSC account where their ownership isn't verified
• Started working on a sitemap parser that will run when the user first signs up, so that I can show some data about that

5-8pm​

• Recursively loaded the sitemaps to get the list of URLs
• Created a db schema to store sites, sitemaps, URLs and the stuff I want to track about them
• Created a UI which lets you push a button to scan for unindexed URLs, which then display below the form for submission for indexing

So what is actually missing for me to achieve my goal:

• Add the feature that auto-indexes your URLs on the paid plan - or drop that for now and just have more submissions in the paid plan for v1
• Try to sell this to someone I know
• Set up a paid advertising campaign

9-10pm​

Transactional work.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (22 / 20)
• Get first paying customer for IndexFast

5-7am​

• Improved the UI more to include the decrementing counter of remaining submissions
• Set up a new Google Ads account and created one campaign - now it's in review
• Improved the UI some more to stream in results as it scans for unindexed URLs
• Fixed a bug with duplicate URLs showing in the result list

5-8pm​

• Remove 'Google' from the ad text because it prevented the ad from showing
• Fixed the broken production deployment
• Created a favicon
• Tested the flow on production and got a timeout error - Vercel employs some limits and I'm still on the hobby plan
• Upgraded Vercel plan to Pro
• Improved the loading behavior on the initial load
• Increased timeout on main page
• Fixed the auth flow for redirects, expired tokens, etc.
• Tested the flow on production again and got an access error on the upgrade page, which doesn't happen on local, and a different error on the billing page, which does happen on local
• Fixed the billing page error and deployed, and then the other page also started working for some reason
• Google Ads started spending
• Checked Google Cloud and my OAuth app is approved
• Attempted to subscribe to my own SaaS service for a production test
• It actually worked and my plan got automatically upgraded from 'starter' to 'basic'!

Some tasks to consider for tomorrow:

• Add the 'auto-index' feature as a button that leads to the upgrade page
• Handle running out of submissions, maybe pop-up an upgrade dialog
• Tutorial upon first opening the main page
• Separate campaign targeting agencies, rather than site owners
• Add more contact information to make the company more legitimate
• Add case studies to landing page
• Run this on my own Shopify store and see what effect it has
• Adjust ad targeting, keywords, spend based on today's data

9-10pm​

Transactional work.

 

[Bounce Back]

Nice! Was the first customer someone you know (like you were planning)?

 

[GoldFibre]

Nice! Was the first customer someone you know (like you were planning)?

I don’t have the first customer yet. It was just me testing the flow on production.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Bounce Back]

I don’t have the first customer yet. It was just me testing the flow on production.

Sorry I misread the goal part as "I got"!

 

[GoldFibre]

​

Weekly Goals​

• Work a total of 20 hours (29 / 20)
• Get first paying customer for IndexFast

5-7am​

• Checked Clerk and I got my first registration
• Checked Stripe and there was no payment, but the person did look at the payment page
• Checked Google Ads and most of the spend is in India and Pakistan - the traffic is cheaper here but I think they are less likely to pay

• Created a duplicate campaign only targeting USA and will see how each one does
• Made an auto-index feature that is paid-only and redirects to the upgrade page if you are on the free plan. UI is pretty good but can be improved

• While connecting my SQL GUI to the Vercel DB I saw that I had two new users that both had proper sites with sitemaps and URLs now in my DB. I suspect that there is a bug that will prevent them from submitting URLs because of the format of their domains. Based on their email addresses and the sites they were managing, they look like they are freelancers doing SEO work for some small businesses
• Found an error in the logs and it was related to authentication
• Added more logging to that function so I can debug
• Activated and set up Vercel analytics to see if it offers anything over GA
• Set up GA and connected it to Google Ads
• Added business information to improve my ads

4-8pm​

• Switched to the official next library for GTM
• Added a conversion event on payment completion
• Fixed a bug with the loading circle prior to Stripe loading
• Fixed the redirect flow after processing the payment, got the conversion event working properly
• Added a conversion event on sign up
• Stopped to think about what is most important to work on next - went through the flow with a critical eye
• Fixed another loading circle bug
• Fixed logic on sequencing of URLs and prioritizing ones that haven't been scanned/submitted
• Supported some bulk inspection of URLs in certain cases, improving the performance

9-10pm​

Transactional work.