Startups usually focus on building their product first, and don’t even try to generate leads or interest until an MVP is ready. Some will just launch on ProductHunt hoping for some lead flow. But in the best case, they will get eyeballs for two days, and then no more leads come in. And building a lead flow is not that easy.
We previously wrote about how we went from idea to our first clients in 6 weeks, but to do that we also needed to build the beginning of a lead flow fast - and without spending too much time on it, as we had a product to build. The only way to make an impact quickly is through outreach; inbound is super valuable but takes more time to build and to get.
This article is about describing how we did it.
A bit of context about us
I'm the Co-Founder at Dataline.io.
Dataline is a proxy that unblocks your analytics, attribution, and marketing tools for your ad-blocked traffic. We don’t re-enable ads or cross-site tracking though, as it’s essential for us to preserve your user’s privacy. We just level the playing field between companies that host their own analytics (including Brave, by the way) and the 95% of companies that can’t afford to. In the end, this means better attribution data, better ad performances, and better lead conversion.
So we sell to marketing teams. And yes, Covid-19 is definitely impacting marketing budgets overall and drastically. But that didn’t prevent us from building a lead flow.
1. Understand who your ICPs and personas are
Our ideal customer profiles (ICPs) are companies that have marketing budgets higher than $50k per month, or companies with a very tech-savvy audience (more prone to use adblockers).
Our personas are:
- data-driven marketing teams - keywords include lifecycle marketing, performance marketing, lead generation, user acquisition, growth marketing
- product / data teams at companies addressing tech-savvy audiences (developer tools, for instance)
- up to the CMO in mid-market companies
Why is this so important? Understanding the keywords that define who you want to reach out to is the first step in discovering and automating an outreach campaign.
2. Choose the right database - in our case, Crunchbase
Depending on the criteria that define your ICP, you will need to use different databases.
In our case, potential public information that could help us segment our audience is: monthly traffic, amount raised (to guess a ballpark marketing budget), industry (to guess their audience). Most of the time, your exact criteria is not a piece of publicly available information, so you need to find some proxy metrics that are public.
Crunchbase had the company database we were looking for with these exact criteria. So we leveraged Crunchbase to build a spreadsheet with:
- Names of companies
- Website URLs
- LinkedIn URLs (very important)
- Total amount raised
- Monthly traffic
3. (Optional) Add the extra information you need
One critical piece of information we were missing was which tools those websites were using and being blocked by ad-blockers. Dataline’s proxy needs to integrate with every single tool to unblock each one, so it would be smarter to only reach out to companies whose tools we unblock.
So we built a simple script that would get the list of blocked tools from a list of website URLs. Thankfully, we already had the list of URLs. The script only took two hours to write.
You might have to do the same, but try to only consider critical information that is actually part of your criteria.
4. Build your list of profiles to reach out - Phantombuster and LinkedIn
Now that we had the list of companies, and the keywords that defined our personas within these companies, we needed to scrape LinkedIn to get a list of profiles to reach out.
To do that, there is this awesome tool called Phantombuster that enables you to do that without coding. Here are the steps to perform:
1. Use their phantom “LinkedIn Company Info” to get the LinkedIn ID for all your companies.
2. Build a list of advanced searches from the LinkedIn company IDs and the persona keywords. In a spreadsheet, they would look like this:
=concatenate(“https://www.linkedin.com/search/results/people/?facetCurrentCompany=[”,Comany LinkedIn ID,”]&origin=FACETED_SEARCH&title=marketing%20operations%20OR%20performance%20marketing%20OR%20growth%20OR%20lead%20generation%20OR%20lifecycle%20marketing”
You put the URLs in the first column without a header, like this.
You put the URLs in the first column without a header, like this.
3. Use their phantom “LinkedIn Search Export” to extract all the search results from your spreadsheet list of search requests.
Now you have a list of profiles with their LinkedIn profile URLs, along with all their public information on LinkedIn. Note that you cannot have access to their email addresses registered on LinkedIn unless they become a connection on LinkedIn.
I would personally advise you to have a LinkedIn premium account at this point. Otherwise, LinkedIn will limit your search results and you won’t be able to get a list of profiles. They offer 30-day free trials, so don’t hesitate to leverage that.
5. Decide on the channel and prepare several campaigns
With a LinkedIn ID, there are many things you can do. For instance, you have their first name, last name and current company. With this information, you can use Phantombuster again to get their Facebook ID or their Twitter ID.
But the smart thing to do is to understand which platforms are the best for reaching out to your audience. In our case, we thought it was both LinkedIn and Twitter.
Then, you need to segment your list of leads into several campaigns. We decided to do that per tool supported.
Why is it important to have several campaigns? To help you iterate, learn and improve your campaigns.
How many campaigns should you do? It depends on the total number of leads you have. On our side, we had between 500 and 1,000 leads for each campaign. The goal is that the number should be high enough for you to get some performance statistics on which you can iterate. Especially for your first campaigns, as you will identify lots of low-hanging fruits to improve them.
6. Automated cross-channel outreach campaigns - LaGrowthMachine
Here is another amazing tool - LaGrowthMachine. It enables you to create automated cross-channel campaigns with just the LinkedIn profile URLs, from which they automatically enrich the lead’s professional email, personal cellphone and Twitter account.
So we created three different sequences focusing on different channels, in order to understand which channels work best. Here is a short description of our sequences:
- LinkedIn sequence with 4 LinkedIn messages
- LinkedIn sequence with 4 LinkedIn messages, followed by 4 emails
- Twitter warm-up (follow, and favorite a tweet, before launching sequence 2)
Here is what our sequence 3 looks like:
Seem complicated? LaGrowthMachine has great templates of sequences to start from, so you won’t start from scratch. That’s what we did, so don’t shy away from using them! Note that if you want to create yours from scratch, LaGrowthMachine’s team have dedicated people to help you build them. And they respect the limits set by all platforms (you can’t just send 1,000 connection requests on LinkedIn in the same day). I would seriously advise you to have a premium LinkedIn account to increase your limits.
7. Iterate on sequence and messaging
Now that we have our sequences and our list of leads, it’s time to launch your first campaign with LaGrowthMachine. The good thing about it is that you don’t need to do anything; it is automated. You only need to reply and continue the conversation once you get a reply from a lead.
There are several metrics to track:
- Connection request accepted / sent: depending on the campaign, I could reach up to 23%
- Reply / connection request sent: this metric can be compared to the standard reply rate for outbound campaigns. The best I could do was 12%, but my average is more like 7%
Several tips!
- If you want to send more invitations because you feel restricted by the daily restrictions of the different platforms (LinkedIn typically allows 150 invitations per day with a premium account), you can create a second identity. That’s why I have two LinkedIn profiles: John Lafleur and Jean Lafleur (I use them both equally).
- You will find that the performance of your campaigns depends on the quality of your audience segmentation and how precise you were in defining the roles you were looking for.
- Iterate the hell out of your messaging. Your copy will always be too sales-y at first.
- You can personalize your messaging using some custom attributes with LaGrowthMachine. For instance, when uploading our list of leads for each campaign, the first column would be the LinkedIn Profile URL and the second column the tools used by the profile’s company. This enabled us to mention those tools in our text, which increased our reply rate.
- If you have great marketing content, LaGrowthMachine also allows you to plan retargeting audiences specifically on the leads you’ve reach out to on LinkedIn or Facebook’s Ads platform, which is great to increase brand trust while also reaching out directly. We haven’t used it yet but definitely a room for improvement by printing blog posts, current clients’ use-cases, incoming Webinar etc... Anything that can help support the sequence by building brand trust.
Our End Results - 10 demos every week
It’s good to set yourself some OKRs. But these are not the metrics mentioned above.
If you are focused on building your product, you can’t spend a lot of time tracking every metric.
We needed a very simple OKR that we could easily track. For us, it was the number of demos per week. Having replies doesn’t imply that you convert them all into demos, and an outreach campaign is only effective if you get leads interested enough to have a demo. But having demos with leads that are not qualified to buy your product - understand having the Budget, Authority, Need, the right Timeline (BANT) - is useless, too. That’s why marketing teams usually have KPIs that are Sales-Qualified Leads (SQLs). But at our stage, we were happy to get demos, as we were more focused towards getting feedback.
In the end, after a month we managed to reach our goal of getting about 10 demos per week.
Let us know in the comments if you have other ideas about how to build awareness fast when you don’t have much time to spend on it!