5 Mind Blowing Automations With Phantombuster, Browse AI, and Clay (Step-by-Step Guide)
Save 1,000s of hours with automated workflows
John Sutter, a Swiss immigrant, established Sutter’s Mill in California’s Sacramento Valley with the intention of building an agricultural empire. He hired an American, James Marshall, to help oversee its construction. In January 1848, Marshall was inspecting the mill when he noticed shiny flakes in the riverbed and quickly realized they were gold. Marshall and Sutter tried to keep their discovery a secret, fearing that a rush of miners would disrupt their operations and plans, but it was to no avail.
Word started spreading around California by March 1848 and when President James Polk confirmed the discovery in a December 1848 address to Congress, it fueled an international frenzy. California witnessed a massive influx of speculators from across the globe, which spurred the development of cities, infrastructure, and new businesses to serve the growing population. While a few prospectors struck it big, most of those who made substantial wealth capitalized on the needs of the miners, such as Levi Strauss, who sold durable workwear, and Samuel Brannan, who ran stores supplying miner. The rush gradually waned in the mid-1850s as easy to find gold became scarce, but it left a lasting legacy on California's development and the American West.
In many ways, the current hype and frenzy with AI is reminiscent of the gold rush. There are a lot of money and resources being plowed into the space, there is a growing awareness that it won’t last forever and a few key players will reap most of the profits, and the biggest winners thus far, such as Nvidia, have been those that capitalize on the needs of the new entrants. However, for the average person, who wants to share in the upside, there is little guidance or direction. Sure, Chat-GPT is impressive and even non-technical people use it for tasks like brainstorming or copywriting, but there are so many more outstanding use cases that most people do not consider.
While not entirely AI based, three of my favorite automation tools are Phantombuster, Browse AI, and Clay. Each can do amazing things, but to do so, it is essential to take the raw features they give you and combine them in such a way that they easily solve common business problems. In 1848, just giving the average prospector a shovel and map and saying, “There are millions of dollars to be made, go find the gold!” would not meaningfully assist them. In 2024, telling someone about AI and saying, “You can save tons of hours and make tons of money with these tools, go for it!” is about just as useful.
In this article, I will be going into step-by-step detail on five very powerful automations I have found using a combination of Phantombuster, Browse AI and Clay. I would like to emphasize that these automations have the potential to be used for either positive or negative means. They could help you to build up your social media and sales lists with great prospects who will love your product, or they could be used to harass and spam people and overrun the internet with bots. On the record, I would like to emphasize that I only approve of using these automations ethically; I do not endorse spamming people or harassing them with high pressure sales tactics. Finally, even if you do not intend to use these tactics yourself, this article should at least make you aware of what is possible and help you understand why you see certain actions occurring online. Let’s dive in.
#1 - Quickly Grow Your Instagram Following with Relevant Accounts
Many people think that if they just post great content consistently on social media, their accounts will grow like gangbusters. In my experience, while posting great content is essential, it is not enough. If you only have 10 Instagram followers, you are facing an uphill battle and are going to have a hard time really seeing much traction. On the flip-side, if you are able to build up a community 1,000+ followers in your niche, you will get significantly more eyeballs and engagement and have a much better chance of going viral.
A reasonable question to ask is, “Why implement such a complex process instead of just purchasing followers?” There are a few reasons buying followers is a far worse option:
-First, buying Instagram followers is obvious and transparent. It almost always involves your account having a history of low user growth followed by a massive growth spike. If someone were to analyze your account using a service like Social Blade, it would be obvious you had purchased followers and you would lose credibility.
-Second, followers you purchase are going to be a lot of spammy and fake profiles and are not going to give your content much, if any, engagement.
-Third, you are giving the Instagram algorithm bad signals about the types of users who are going to be a good fit for your content. This is going to make it much harder for you to get your content in front of the right users in the future.
-Fourth, if you do eventually try to monetize or offer something to your audience, you will get zero interest from your fake audience.
Without further ado, let me walk you through the process to substantially grow your following with the right types of accounts:
1). Use the site All Hashtag to find relevant hashtags for your niche. You can just start with a single phrase and it will give you a list of all of the hashtags similar to it.
As an optional step, if you want a few more hashtags, insert each of these one at a time into the IG search bar and see if any other good hashtags come up as a result. For example, you might search ‘#b2btips’ and find ‘b2badvice’ as a relevant hashtag.
2). Once you have a full spreadsheet with hashtags, it is time to use the Instagram Hashtag Search Export phantom. This enables you to extract the most popular Instagram posts linked to a set of hashtags. It will also tell you the account behind every post.
3). Once you have this list of top posts, you can use the Instagram Post Commenters Export to export a list of all of the people who have commented on these posts. Why start with commenters? It is because commenting on a post shows a very high level of engagement. As this resource from Statista shows, on average in 2024, most Instagram posts get 33x the number of ‘likes’ as ‘comments’. For that reason, users who comment have a history of high engagement and are thus likely to highly engage with your brand’s content as well.
4). Purely from steps #1 - #3, you should have built up a massive list of relevant accounts, but if you want to take it a step further, you can use the Instagram Follower Collector phantom and the Instagram Photo Likers phantom to expand your list further.
5). Once you are confident in your full list, we are going to engage with these accounts. Based on past tests I have run, the steps of ‘follow’, ‘like’, ‘like’, ‘like’ tend to be a really successful way to engage accounts and get them interested in following you back. The first step is following the accounts in our list, which we can automate with the Instagram Auto Follow phantom.
6). Finally, we are going to ‘like’ 3 posts from each account. There are two phantoms needed for this: the Instagram Profile Post Extractor phantom, which allows you to extract posts from a list of Instagram accounts, and the Instagram Auto Liker phantom, which allows you to ‘like’ a series of posts from each account.
If you have done this properly, you can expect somewhere between 10% - 20% of accounts to follow you back and you can expect to see strong interest and engagement in your profile from these accounts. While a 10% - 20% follow-back rate is not insanely high, if you follow steps 1 - 3, you should be able to build an initial list of thousands of accounts to pursue. If you add in step 4, you can easily scale this to tens of thousands of accounts. Once you get to numbers this large, 10% - 20% follow back and engagement is more than enough to quickly scale up and build a huge IG account.
While automating this strategy initially is a great way to engage accounts at scale and grow your own number of followers, I highly recommend that you continuously engage with your new follower’s content over time on your feed and show genuine interest. The intent is not to game the system and benefit at other user’s expense. As a final note, please be mindful of rate limits. If you use Phantombuster automations beyond the recommended limits, your social accounts may get flagged.
#2 - Grow Your Number of LinkedIn Company Page Followers
I have used this exact process to help multiple companies to quickly grow their company pages by thousands of followers. If you have your own business and you have a team of leaders and executives, I would highly suggest you each run this process to maximize reach and grow things as quickly as possible.
The best way to grow your number of LinkedIn company page followers is by leveraging your existing network, your LinkedIn connections. But, what if you do not currently have a large network? Let’s start there:
1). If you want to expand your network, the first step I would suggest is finding people you know through your personal or professional lives but who you have not yet connected with on LinkedIn. Going to the ‘My Network’ tab on LinkedIn and checking the ‘People You May Know’ section is a great place to start. This should easily uncover 50+ people (potentially way more) who you know on some level and could easily connect with.
2). After exhausting the ‘People You May Know’, who tend to be very close connections, it is best to then connect with people who attended your school or worked in the same companies. You can use either the Sales Navigator Search to Lead Outreach phantom or the LinkedIn Search to Lead Outreach phantom to find them.
For myself, I went to Darien High School, then Cornell, did a summer program at the London School of Economics (LSE), worked at Vivaki and MightyHive, did a coding bootcamp at Hack Reactor, and then started my own company Stackmatix. While schools as big as Darien High School, Cornell, or LSE are too big to try to connect with all of their alumni, a good place to start is to apply some filters to find people who were in my same year, major, or in the case of college, same fraternity as me.
For the workplaces, assuming your company is 250 people or less, it is not unreasonable to find and connect with the entire company via LinkedIn. If you work for a much bigger corporation, I would suggest applying a few filters to find people who work in the same unit or division as you. If you want to take things a step further, you can build additional lists of people at other companies who do similar types of work. For myself, I might do searches for people whose job title includes ‘Growth’, Go To Market’, or ‘Startup Advisor’ and send them connection requests.
If you have followed this approach and used the phantoms to send out connection requests, you should easily be able to expand your network by 300 - 500 connections (or potentially a lot more).
3). Once you are confident in the size of your LinkedIn network, the next step is to export your list of connections using the LinkedIn Connections Export phantom. Of the connections you have, there will likely be a few you do not want to reach out to (e.g. maybe there are a few people you have a poor relationship with), or maybe it is just the case that you only want to reach out to contacts to whom you have been connected for 6+ months. Either way, by exporting your list of connections first, you can curate the list you want to reach out to and make sure you have control over the process.
4). With your list ready to go, you can use the LinkedIn Message Sender phantom to individually message each of your connections. I have found from experience that a simple message works best:
“Hi [First Name],
My company _____ does _____ and we are making a push right now to grow the following of our LinkedIn company page. Would you be open to giving us a follow [link here]? I would really appreciate it and would be happy to return the favor or help you in some other way.
Best,
[Your Name]”
There are a few caveats to this:
-Just send a single message. Do not do multiple follow-ups - this borders on spam and is not fair to your connections.
-If any of your connections ask for something in return, make sure you follow through and go to bat for them. It is essential that you take this seriously and do not try to be a mooch.
-Finally, while LinkedIn also has the LinkedIn Company Page Inviter phantom to grow your page, I do not think it is as effective. First, it requires credits, which you have a limited number of. Second, it does not allow for the extent of personalization with your message (including offering something in return).
#3 - Building High Quality Sales and Marketing Lists
Let’s say that you want to do a sales or email outreach campaign, but first need to build up a solid list of prospects. As I state in this article, while you can purchase outreach lists, high quality lists tend to be pretty expensive (usually $3 - $5 / contact). For that reason, if you can build your own custom list, it can go a long way and save you a lot of money. As far as how to do this, I am going to show two different scenarios:
-You have a list of Instagram profiles and nothing else.
-You have a list of LinkedIn profiles and nothing else.
Instagram Profiles
A great way to build a list of niche relevant Instagram accounts is to use Clay’s AI agent, Claygent as described here. While your initial plan might be to direct message (DM) these contacts on Instagram, I have found from past experience that this is largely ineffective. In the event that you are not connected to the accounts you want to reach out to, your message will be placed in their ‘Requests’ folder and is almost guaranteed to be ignored. Even if you are able to get into their inbox, many influencers have turned ‘Read Receipts’ off, meaning that you have no way of knowing if they have even seen your message. For that reason, email outreach is a far better way to get seen. Here is a step-by-step process to get someone’s email with just the Instagram profile:
1). You can scrape all of the available data from each Instagram profile using the Instagram Profile Scraper phantom.
I did a small test on about 150 profiles and of these, it was able to find emails for 112 of them, a solid match rate of 75%.
2). For any Instagram profiles you are unable to scrape, I would suggest using Clay’s AI agent, Claygent, to run a few searches and see if it can find the remaining emails. Here is the prompt I used to do that. This prompt enabled me to find emails for an additional 14 contacts, bringing the overall match rate to 86%. If you do not have Clay, Chat-GPT is an okay alternative, but you will have to do this one at a time and it will be a bit of a hassle.
LinkedIn Profiles
1). To find a list of relevant LinkedIn contacts to go after, you can use either the LinkedIn Search Export phantom or the Sales Navigator Search Export phantom. Just apply a few filters to help narrow down your ideal client profile (ICP) and you can quickly build a list of thousands of relevant prospects.
2). I would then suggest uploading this data into Clay. You can use their data enrichment to find ‘Work Email’ and ‘Personal Email’ based off the LinkedIn profile. Their process does a data match against 10 different data providers, creating a very high likelihood that you will get a match (from past tests, I would estimate a match rate of 95% or higher).
It is worth noting that in the past, in the LinkedIn platform, you were able to export your entire list of LinkedIn connections to a CSV along with their emails. While the feature does still exist to export your LinkedIn connections, you will only be able to see the emails for a small percent of them. When I tried it myself, I only got a match rate of about 2%. Consequently, Clay’s ability to find work and personal emails is more valuable than ever.
#4 - Keeping a Pulse on the Market and Plugging Into Relevant Conversations
In my previous article, I talked about the the importance of online reputation management. While the primary focus of the article was giving you a clear checklist to help you to identify your brand’s current reputation and where it could be improved, my final suggestion was to set up a reputation monitoring system. Setting up Google Alerts is a fantastic place to start for this, but it is not enough. Conversations that are important for your brand are happening all over the internet on a daily basis and you are going to miss out on them if you fail to track them.
Furthermore, a good reputation monitoring system not only allows you to play defense (e.g. jumping in when your brand is mentioned negatively), but it also enables you to take proactive measures to enhance your reputation (e.g. finding conversations in your business’ area of interest and jumping in to offer value). With Browse AI’s pre-built robots, you now have the ability to build a robust reputation management system. Here is how:
1). The Extract posts from Reddit search robot enables you to extract relevant posts from a Reddit search into a spreadsheet. Considering that Reddit is used heavily by Millennials and Gen-Z, and considering that many of the conversations that happen on Reddit are product reviews or users asking for feedback about a product, this robot is a great way to monitor if your brand gets mentioned, see if other areas of interest are discussed, and engage proactively.
2). This Monitor Country Trends on Google Trends robot will notify you when a new trend is emerging on Google Trends in a selected country (or automatically update a spreadsheet with this data). This is useful to be able to see where the market is going and get new ideas for your content strategy.
3). Hacker News is a very popular online news board with heavy usage among tech and startup people. The Extract News Items by Keyword from Hacker News robot allows you to monitor news items by keyword. Hacker News articles feature a plethora of comments and have highly engaged users, so if you can jump into crucial conversations and provide valuable insights in a timely manner, you are sure to get a lot of eyeballs and engagement.
#5 - Competitive Price Monitoring
Let’s say that you own a popular online e-commerce brand with over 100 products. While you have plenty of loyal customers, you are in a price sensitive industry, and if you do not keep on top of competitor’s price adjustments, you are likely to lose a lot of sales. This is especially true during busy and time sensitive periods like Black Friday and Cyber Monday. How best can you monitor competitor’s prices and ensure you are staying on top of the competition? Browse AI’s robots are the answer.
With Browse AI, you can build a custom AI web scraper to check the prices for your competitors each day and insert them into a spreadsheet. Once the data has been updated each day, you can use a tool like Zapier to alert you. You can also create formulas in your sheet to flag any large price fluctuations. While keeping track of hundreds of competitor’s prices would have been a herculean task in the past, with Browse AI, it becomes easily doable.
What does a custom AI web scraper look like? Check out the video below. While in the past, you needed to understand coding and all the nuances of a site’s structure to create a web scraper, today, you can just click on the items on a site you want to scrape and the AI will do the rest.
Conclusion
While the AI revolution is marching forward at an increasingly accelerated clip, AI will only prove useful to people to the extent that it solves real world problems. There are so many amazing AI tools and solutions out there, but without the proper knowledge on how to apply them, they are like the untapped gold hiding in the California riverbeds. I hope this article provides a glimpse into what is possible with AI and automation and gives you the motivation to begin developing your own use cases.
Incredible what you can do with those tools, thanks for sharing!
Definitely going to test these out