I have experience with this type of customer acquisition. It works, but it's painful in that it is extremely boring.
To take some of the boredom out, I did two things: (1) Used oDesk to gather email addresses with business information and (2) Built a custom system to send out individualized emails.
(1) My experience is similar to Tracy's in that the quality you get from Elance/oDesk will be poor . . . at first. The reason why this is so important is that if you want to do as Tracy describes, which is send personalized, non-spammy looking, friendly emails, you have to get business data (company, CEO's name, etc.) free of any errors.
I found freelancers on oDesk and get them the Inc 5000 and told them to get me the email address of the CEO, the company name written in plain English as you might use it in a sentence (ie, not Hacker News Incorporated, but Hacker News), the CEO's first name that he/she goes by (e.g., Katie, not Katherine), etc. I had to go through it by hand like a teacher and point out every mistake made, and I did this about 3 times before I started to receive higher quality data.
Then got those people I trained to train others and I held the "managers" responsible for any errors. I hired mostly native English speakers from the Philippines.
(2) I had the freelancers enter data into a shared spreadsheet and wrote a script to dump that into a database which would prepopulate templates. I would then inspect each email and send them out one by one (this step could perhaps be skipped later on).
Each company had a "thesis", e.g., Inc 5000 companies want to use our product because they're growing quickly, and various other attributes that would determine which template was sent for which company and then compiled analytics accordingly (A/B testing different templates for a given thesis was built in).
The system then scheduled follow ups after a given amount of time up to 3 follow ups before stopping.
This stuff works, but it's a grind. If you start to pay freelancers for the info gathering, you should carefully watch your expenses and track the relevant stats (conversion rate, CLTV, etc) to make sure that you're actually making money.
One last thought, I'm sure Tracy's post and my comment will get pushback about being spammy, and I sympathize. Even patio11 got pushback for his post encouraging people to use email more as a channel. I think it's very important to craft an email that's meaningful and delivers some value to the recipient, and that means targeting and craft in the email itself which was at the center of my efforts and you can see the same in Tracy.
The Email sample, with simple and clear justifications, has to be the most valuable point in the post. I'm consistently amazed by writing that seems so genuine that I knowingly ignore the fact that it's spam.
This is excellent. I have a list of possible first 50 potential customers for something that I am about to start soon and was thinking along similar lines. In my case however, I know about 30% of the 50 potential customers as well.
I strongly think that hand crafting these first emails will be very important. I know some people are talking about automations through crawlers,twitter etc. but for first few initial customers, I would go with the way mentioned in this blog.
Right. It's also important to have good, high quality businesses on the site when it launches to give a really awesome impression to potential users and new customers. Emailing semi-personal emails really doesn't take that much time.
Very useful post. It's great to see these kinds of processes documented, might be nice to see some funnel metrics too - talking to some folks that also do (or need) small business customer acquisition, it seems there are a lot of sticky points in that process. My personal experience with small businesses is fairly hands-off, and from a large company, so it's nice to see the other side.
Couple of thoughts:
I'd expect that having successful vendor directories under your belt makes it more likely that someone will sign up, because you seem way more legitimate. Perhaps there are some stats from the most successful site you could use in the sales technique, e.g. "Our most successful invitation vendors see over 10 new customer leads a day from our site". (Or perhaps that's something for later on in the process, if a customer isn't convinced.)
Love the human aspect, the number of emails I (still) get at my old company address from people who clearly have no idea what the company does or what I would be interested in... It's pretty astounding.
I'm torn on the "I'll create your profile for you". I've had a few spammy emails where they effectively said "You've been listed in $directory! Here's your listing! Now pay us so you can do stuff with it". Urgh. Red flags. I wonder if there's some intermediate step that works even better, like "I can create a profile for you, here's the information I have about you" - a little more personalisation.
It also just occurred to me that to really get inside the customer's head for something like this, you could set up a vendor site yourself and see what kinds of emails come in :) Probably more useful if this customer acquisition process is going very badly, than immediately at the start.
"I'd expect that having successful vendor directories under your belt makes it more likely that someone will sign up, because you seem way more legitimate. Perhaps there are some stats from the most successful site you could use in the sales technique, e.g. "Our most successful invitation vendors see over 10 new customer leads a day from our site". (Or perhaps that's something for later on in the process, if a customer isn't convinced.)"
This is the same process I used for the first one I launched, didn't see many differences in response rate. And great idea on the leads, that's a good point.
"I'm torn on the "I'll create your profile for you". I've had a few spammy emails where they effectively said "You've been listed in $directory! Here's your listing! Now pay us so you can do stuff with it". Urgh. Red flags. I wonder if there's some intermediate step that works even better, like "I can create a profile for you, here's the information I have about you" - a little more personalisation."
Well, I'm definitely trying to not do that, which is why I'm asking permission to create the profile for them, not creating it beforehand. If they agree, I'll create it on my end (not live), take a screenshot, and then send to them for review. Agreed on the red flags. :)
limedaring - have you tried automating any of these steps? we've had a lot of success with using crawlers to find customers -> Amazon Turk to gather contact info -> templated emails.
True! You have to be careful with the quality (the big issue that I had using Elance). I've done something very similar crawling Twitter lists, and probably could work with crawling Yelp. I'll see about adding it to the article. Thanks!
To take some of the boredom out, I did two things: (1) Used oDesk to gather email addresses with business information and (2) Built a custom system to send out individualized emails.
(1) My experience is similar to Tracy's in that the quality you get from Elance/oDesk will be poor . . . at first. The reason why this is so important is that if you want to do as Tracy describes, which is send personalized, non-spammy looking, friendly emails, you have to get business data (company, CEO's name, etc.) free of any errors.
I found freelancers on oDesk and get them the Inc 5000 and told them to get me the email address of the CEO, the company name written in plain English as you might use it in a sentence (ie, not Hacker News Incorporated, but Hacker News), the CEO's first name that he/she goes by (e.g., Katie, not Katherine), etc. I had to go through it by hand like a teacher and point out every mistake made, and I did this about 3 times before I started to receive higher quality data.
Then got those people I trained to train others and I held the "managers" responsible for any errors. I hired mostly native English speakers from the Philippines.
(2) I had the freelancers enter data into a shared spreadsheet and wrote a script to dump that into a database which would prepopulate templates. I would then inspect each email and send them out one by one (this step could perhaps be skipped later on).
Each company had a "thesis", e.g., Inc 5000 companies want to use our product because they're growing quickly, and various other attributes that would determine which template was sent for which company and then compiled analytics accordingly (A/B testing different templates for a given thesis was built in).
The system then scheduled follow ups after a given amount of time up to 3 follow ups before stopping.
This stuff works, but it's a grind. If you start to pay freelancers for the info gathering, you should carefully watch your expenses and track the relevant stats (conversion rate, CLTV, etc) to make sure that you're actually making money.
One last thought, I'm sure Tracy's post and my comment will get pushback about being spammy, and I sympathize. Even patio11 got pushback for his post encouraging people to use email more as a channel. I think it's very important to craft an email that's meaningful and delivers some value to the recipient, and that means targeting and craft in the email itself which was at the center of my efforts and you can see the same in Tracy.