We implemented an automated marketing campaign and replaced Hubspot with a mix of 10 individual tools and a handmade process, reducing the subscription cost by 90%. Here’s how it worked.

We planned a campaign for a — think Patreon with cents — that has both a Business to Business and Business to Customer element. In order to market and manage thousands of leads, in our case authors and content consumers, we needed a streamlined automation of both inbound and outbound activities.This is known in the industry as marketing automation. And the pinnacle of programs to do this with is HubSpot. Its Patrick Mahomes, Simone Biles or Megan Rapinoe all rolled into one super-performing software suite. It can do everything and then some. It is more than a software, they have a mini university attached that offers free courses in marketing. You even get certificates for finishing these courses. They don’t really mean anything, but you feel good. Kind of. They give you a personal onboarding experience and help you get the most out of the sprawling universe of a suite.Everybody loves Hubspot. And if they don’t it’s because they love Hubspot’s nearest competitor Salesforce, a giant in its own right. I would’ve loved Hubspot. But for our marketing behemoth, we would have had to fork out $4,000 per month for the service and the contacts that service made on our behalf. For that amount we could hire a developer or like, 10 human marketers. Just kidding. It was way over our budget.On my resume, I tout myself as a MacGyver — give me a paper clip and a yoghurt and I can construct a self-playing baby grand piano. So, I decided to ply my jerry-rigging skills at marketing automation. Why not try to replace the one-stop-shop Hubspot with a mix of cheaper stand-alone tools and a few tiny workarounds? I am old-fashioned, I rather have one more human in my team than the best software. Will it pay off? This is the story of a journey, tumbling down the rabbit hole, and Kansas going bye bye.

In and out, bound to happen

From the marketing textbook comes the basic strategy:1. Lead generation — collecting and qualifying the right contacts2. Outbound — contacting target groups directly where possible3. Funnel — directing them to the homepage4. Engagement — getting them to converse with us5. Targeted ads — follow them around6. Social & PR & content marketing — creating touchpoints and inbound opportunities7. Conversion — get authors to register8. Retention — get authors to actively use and promote the service9. Upgrade — transform power authors to brand ambassadorsHaving very little funds for B2C marketing, we planned to reach out to content creators as they are the main benefactor of a micro donation service. We’d convince them to support the service and tell their audience: “If you want to give me a few cents as a thank-you, install the app and spread the love.” The authors were our B2B segment which comes with the advantage that we can address them. We decided that our first hub was going to be LinkedIn. Put “Youtuber” in the search and you get thousands of authors. We’d contact them all. That is the outbound part.When they respond and visit the app’s website, we’d need a process to convert them to brand ambassadors, help them tell their audience and convert the fans. That is the inbound part. Naturally, this is supported by targeted ads, social media marketing and public relations.

A visual representation of Pipedrive’s lead qualification workflow (Kanban-style) / by Pipedrive

Lead generation

We planned to contact 30,000 authors per month, showing them the free and fee-less service. We split between LinkedIn and cold mails (for regions where GDPR considerations don’t apply). We created author segments by content topic, media and country. Our “yellow pages” is LinkedIn. We used a tool for extracting E-Mail addresses from LI users: ($200 / 50,000 emails = $70 / month)Skrapp does verify email addresses but we used an additional tool to cross-check: ($24 / 10,000 emails = $36 / month)The heart and soul of automated marketing campaigns is the CRM, the database where all leads are stored, status updates are added, actions depending on rules are set. This is the tricky part in the stack as the best ones are expensive and the cheap ones lack in functionality as well as comfort. We needed an app that can serve unlimited contacts without the heavy price. After trying six or seven tools, we finally settled for: ($49 / month)This is not a full-blown CRM, but a sales-pipeline tool. It is a bit cumbersome to use but it has all the functionality that we needed and didn’t have from other tools. We manually inserted csv files of the generated leads. Since this process isn’t done on a daily basis, we went without an automatic data connection via an api-tool.This process is not ideal, as many that follow. Manual data export and import costs time and comes with an additional error vector.

Stats-view in Quickmail / by Quickmail

Outbound

In order to avoid spam, LinkedIn limits the amounts of connections one can make and punishes those who do. The company also bans the use of external automation tools for personalized connection and engagement purposes. We tried out the tool: but eventually decided to have a human do 100–150 personalized connections each for three of our profiles per day. This is not verboten and allows for deeper personalization.We tested numerous cold-email clients and chose: ($49 / month for one address plus $29 for each following = $109 / month for us)For coldmailing, we didn’t want to use a drip campaign (having a script of 3–5 different emails in a status-related process). One email and if there is no response, then this lead is left in peace.Apart from the price, the single most important factor for a mailing app is deliverability. What you want to avoid is for your domain to be registered as spam. There are several things to do to prevent that. We had booked a private workshop with a cold mailing professional which is a worthwhile expense, as the consequences of a bad campaign can be lethal for the project. Quickmail was also the expert’s choice especially in regards to deliverability.Again, we set up a csv exchange between Pipedrive and Quickmail. We’d export today’s segment leads from Pipedrive (and ticked a status box) and fed it to Quickmail. The latter one is good for sending, but not for subsequent conversations. But Pipedrive can do that which kept the data close to our central hub.

Chatbot example in Drift / by Drift

Inbound / conversion

In order to do rapid development, we chose Wordpress as our main landing page platform. One advantage is that WP can double as a knowledge base tools via a plugin for both users and customer support. We’d switch to a landing page generator once we started with A/B testing. But right from the start we wanted to use conversational marketing in form of a chatbot. We chose: ($400 / month or $50 / month)That is a fantastic tool. You create playbooks for chat conversations that lead to a human response or a calendar for an appointment. It comes with all the metrics and even email integration — the user clicking the link on the mail is then addressed by the chatbot on the website with her name and a specific playbook. We didn’t even plan to incorporate too much bot use as we wanted to converse human to human as fast as possible to get insights as much as conversions. Drift offers quite good books on conversational marketing and the best thing: they offer an extreme discount for startups with less than 10 people. The integration with Pipedrive is tricky and works so-so with an intermediary tool like Zapier. But even without any integration would I always come back to Drift.We decided against dynamic pop-ups for peddling newsletter subscriptions and thus against using the expensive market leader Privy ($140 / month). Let’s test without and see if we can get decent sign-ups. The power of the story instead of the power of being annoying.

Retention

We were very uncreative with the choice of a newsletter tool with: ($15 / month)We didn’t immediately plan drip campaigns since by now we were running thin with human resources. We’d keep that for a later stage. Given that we already used Quickmail, we’d try it for drip as well even though it didn’t look promising during our test-runs.I am not going into details for the social media management, as that wouldn’t be core of a hubspot replacement.

Example Dashboard in Databox / by Databox

Analytics

Tools create statistical meta data like cows create Methane. There is too much, barely visible and it is scattered among many apps. It comes as no surprise that there are meta analysis tools combining all data source (plus analytics, adwords, social channels) and create cross-data reports. With this we can measure impact and improve KPIs. It did surprise though that none had a either a good user experience, or a connection to all the tools we used. We used the least objectionable tool with: 
 ($50 / month)

The experience

If you have followed this far you can imagine that it took ages to find, research and test all the tools and their closest competitors. It took weeks. It took even more time to create a robust process from gathering data to measuring KPIs. That took months. I have to admit that we could not go completely without Excel, especially in lead generation and in managing segments. The results were fine, we exceeded industry averages such as having 40% connection rate, over 20% with good conversations and nearly 10% conversion rate with authors. Having a workflow and tools did matter but far less than the story and the human element of just being nice to people.

The conclusion

Having set up this elaborate tool-stack with hand-made processes felt a bit like a card-house. I was glad to see that no wind blew it over, but it wouldn’t have surprised me. We gathered the data with skrapp and enriched it with queries to various social media sites. We aggregated, cleaned and segmented the data in Excel. At least parts of the complete marketing process had to be modeled in the world’s worst database. That was a side effect of not having one strong data repository as Hubspot would have been.Excel is to data what Powerpoint is to a theatre performance. A nightmare. It’s slow, unreliable and it is so easy to make undetected and undetectable errors.It is possible to channel 1,000 leads through your pipeline on a daily basis for under $400 in monthly subscriptions. But if money wasn’t an issue, I would always use a near-fully integrated tool like Hubspot (or Salesforce or Adobe — whichever floats your marketing-driven boat).

The project

I cannot report on the long-term results for our project as we are not there yet. COVID-19 had a massive impact on us as it did on many others. Midway through the first phase of our campaign we lost our financial support and had to halt and postpone the wide rollout. We haven’t yet bounced back so only time will tell. But that is a story for another time.