Episode 43 | The Customer Acquisition Process

Show Notes


[00:00] Rob: This is Startups for the Rest of Us: Episode 43.

[00:04] [music]

[00:13] Rob: Welcome to Startups for the Rest of Us, the podcast that helps developers be awesome at launching software products, whether you’ve built your first product or you’re just thinking about it.  I’m Rob.

[00:22] Mike: And I’m Mike.

[00:23] Rob: And we’re here to share our experiences to help you avoid the same mistakes we’ve made.  What’s the word this week, Mike?

[00:28] Mike: Well, I just got done, I think it was yesterday or the day before, working through what I’ll call some Entity Framework stupidity, and it’s pretty blatant.  The problem I ran into is that I wrote a bunch of code and I finally got it working, took me forever to do it, because I’m dynamically creating some objects on the fly that are supposed to have underlying meta data so that when I go to sort them there’s information from which to sort it on.

[00:55] And I deployed it onto…I was working on my machine, on my laptop.  Everything’s working fine.  I go to deploy it to this production server and most things work, but there’s certain pieces that don’t.  And I could not, for the life of me, figure out what it was.

[01:09] I finally came to the realization that there’s this configuration flag or something along those lines, and it’s a configuration setting in the XML file that’s called Provider Manifest Token, which defines what the database is on the other end of Entity Framework, which seems odd, because what it does is it basically did a translation when I did a new date/time, and it created it as a date/time 2 on my laptop, which has got SQL 2008.  And on the server I deployed it to, it’s got SQL 2005.

[01:42] So because it was set to 2008, it says, “Oh, well I’m going to create this as a date/time 2 object,” and SQL 2005 has no idea what it’s talking about.

[01:53] Rob: Brutal.  So how long did it take for you to figure that out?

[01:55] Mike: It probably took me two or three hours.

[01:59] Rob: That sucks.

[02:00] Mike: It probably would have taken me a lot less time, but I basically had to compile the code and put it out there, and then it didn’t work.  And because I didn’t have direct access to the server so that I couldn’t hook a debugger to it or anything like that, I had to resort to print statements.  It was a nightmare.

[02:18] Rob: Yeah, debugging on the server…I’ve had this happen before where something works on your dead machine, doesn’t work on the server, and you have really limited resources at that point because it’s like you said, you can’t do the stuff through debugging and you’re just left to the mercy of response, right, which is painful.  So yeah, we’ve had things take, at my old job, we’ve killed eight hours, two of us, trying to troubleshoot something similar to this and found out that it was a similar issue.  It was like a config issue with this one DLL.  Yeah, that sucks.

[02:48] Mike: Yeah.  The thing that really gets me is that every time I make a database change now though is I have to go in there and make sure that I change that back from 2008 to 2005, because it will suck in the value from the database.

[03:01] Rob: Oh, wow.

[03:01] Mike: So every time I do it, I basically have to reset it and then recompile and then I can do my thing; it works fine.

[03:10] Rob: Right.  Oh, man, that’s an argument for, I guess, getting like continuous integration server setup, or I should say just like a deployment script that can do that for you, because otherwise it’s a manual step.  I guess that the worst case you set up a process and remember to do it.  But yeah, that’s a bummer, man.

[03:27] Mike: What about you?

[03:28] Rob: Well, me this week, I’m just starting to work on my talks.  I have several talks over the next couple months.  I’m speaking in Vancouver in a couple weeks and then in Michigan, and then at MicroConf in Vegas—the conference we’re putting on.  I just got asked to speak here in Fresno at a startup competition called 59 Day of Code. So I’m pretty excited about that.

[03:47] I judged there last year and this year they’re doing actual speaking stuff, like keynotes and such.  So I’m excited.  It’s like the “season”.  It seems like all the conferences are kind of in the spring and early summer and they’re all packed together.  And so, I don’t travel much during the year, aside from what I do for fun, and all of a sudden I have four trips, essentially, planned.  Or I should say three trips because Fresno is not really a trip.

[04:10] So yeah, getting moving with that.  I have a lot to say.  It’s been about a year…I guess it’s only been about six or seven months since I last spoke.  And I learned so much from talking at Business of Software last year and just seeing what the really good talks were, and then seeing the less good talks.

[04:27] Mike: So you viewed your video is what you’re saying?

[04:28] Rob: Yeah, exactly!  I threw popcorn at my own video.  No, but I did.  I watched my video and I was like, “OK, this is what I should have done differently.  So it was super helpful to have that.  And I’m excited to kind of take that and step it up a notch and just hopefully ideally get better this time.

[04:47] Mike: Yeah.  You know, I went back and watched a video of Patrick McKenzie’s…was it Lighting Talk?  Or what do they call it?

[04:54] Rob: It was Lighting Talk, yeah.

[04:55] Mike: So I went back and watched his a couple of days ago.  And although it’s entertaining, I have to be honest.  When you start watching it, if it’s the second or probably third time that you’ve watched it, you start to see past the jokes and see some insights of what I would quite frankly call brilliance in his speech.  I mean because it’s very…it’s dead-on.  And I can’t believe that one person could have such great ideas.  I mean, obviously, a lot of people have great ideas.  But it’s just amazing the sheer number of them that he kind of strung together in that little short seven minute segment.

[05:31] Rob: Yeah, I agree.  I think you’re right.  I think the humor did kind of throw up a smoke screen and people laughed, and you just kind of go with it.  But you’re right, there were a lot of undertones and some original thinking in there.

[05:43] I think that’s a challenge at the lighting talks.  I actually kind of get irritated that they’re all so funny, because it makes it harder to learn from, frankly.  Because either I’m laughing or everyone around me’s laughing, and it’s like  you don’t get the real point of it. Maybe that’s the point.  Maybe they’re supposed to be pure entertainment.  But I think you do kind of need to watch them a few times in order to get something out of them.

[06:04] What else?  Anything else going on?

[06:06] Mike: Well, as you know, I’ve got that self imposed deadline that I’ve set for the end of June that I’m working on for AuditShark.  And I kinda set aside, not this coming week, but the following week, the two weeks after that I’ve basically set aside to do nothing but work on the product.  So I kinda shuffled off everything else that I’ve got going on and I’ve allocated about 100 hours worth of time that I can just sit down and do development.

I don’t know if that will be enough to be perfectly honest.  I’ve got a phone call next week with somebody on Microsoft’s architecture teams to kind of help me figure out what some of the next steps are and what they should be for putting the product out in the cloud.  Because obviously, there’s two different options.  One, I could use Azure, or I could kind of build my own Azure-like services.  And I hesitate to think which one’s actually the right way to go for me, because I don’t know the answer.

[06:58] Rob: Yeah, I’ve heard that Azure is really hard to deploy too.  Like even in talking to some MVP’s at the Summit, they were like, “Yeah, that’s where Microsoft just kinda screwed this thing up.”   I’ve used it, but according to them…

[07:12] For any listeners, Azure is basically Microsoft’s Google App Engine.  You know, Google App Engine you can just deploy some Python code and it kind of scales infinitely and you pay per whatever…What is it?  You pay per minute?  You pay for CPU cycle?  I don’t even remember what it is.  But you pay…it’s like a utility; it’s computing as a utility.  And that’s what Azure is, for .NET code.

[07:32] Yeah, I’ve heard that you can’t just…you have to actually create something as an Azure project in Visual Studio.  You can’t just take a normal web app and just convert it or anything.

[07:43] Mike: You can to some degree.  You just have to be sure that the things that you’re doing are kind of with the mindset of Azure…

[07:53] Rob: See, that’s the disappointing part.  I mean I know that with a web app, if you’re going to scale it like that, certainly there has to be some constraints to it.  But from what these guys are explaining to me…and they told me some specifics that I don’t recall now.  It was, what, a month ago now or six weeks ago?  But they were basically saying like you had to do some things that, to me, are just unreasonable to have to think about when you’re building your app.  Because Azure shouldn’t be that much different than just deploying it to a web server.

[08:19] Mike: But it is though.  It really is.

[08:21] Rob: But it shouldn’t be from…as a developer, the experience shouldn’t be that different.  They should get it as close as possible.  And from what it sounds like is it’s not very close at this point.  You really have to do a lot of things differently.

[08:35] Mike: Yeah, I wish you had some very specific…

[08:38] Rob: I do too.  The guy who was doing it, I don’t want to mention his name, but he kinda walked me through what it took for him to get his app up, and it was shocking.

[08:50] Mike: Was it an app he’d already built and then he decided to put it into Azure?

[08:54] Rob: I think so.  Yep.

[08:56] Mike: See, what would be a problem.

[08:58] Rob: But it shouldn’t be.  It shouldn’t be a problem.  The thing is, like with Google App Engine, you have to make some adjustments, but you don’t have to do all this crap.  I mean he had to do hours and hours of work.  And then every time he wanted to redeploy, it was like 20-30 minutes of work just to redeploy his little web app.  And it’s like I can go to my shared hosting and FTP it up in two minutes.  And he’s like, “Oh no, it’s not like that.”  He’s like, “You gotta do all this config stuff every time you redeploy.  I don’t know.  That was a big deal.

[09:27] So if someone’s out there and I’m totally full of crap…I mean again, this is like second or third hand at this point.  I haven’t done it myself.  But there were two guys there who both had done it independently and they said it, so I’m taking it at this point as not fact, but as kind of a reality.

[09:44] But Mike, you know, if you do this, I’d love to hear your experience frankly.

[09:49] Mike: Yeah.  I mean the stuff that I’ve done so far is all using the local development space that they give you.  I mean it’s a cakewalk to work with because it’s really not any different than .NET.  I mean maybe the actual deployment of that stuff is a little different, because I haven’t gone through that stuff yet because I haven’t actually signed up for the service or anything.

[10:05]But if you’re taking an existing application and then porting it over to Azure, I could see how that could be very problematic, just because of the way that you need to think about things like message passing, and scalability, and things like that, you can’t just do things the way you’ve always done them when you’ve got the expectation that everything is on a single box or you’ve got a SQL Server and you’re frontend web app.

[10:31] So, for example, if you take a session…if somebody instantiates a web session to Azure.  Well, between one request and the next, you’re not guaranteed to be on the same server.

[10:42] Rob: Sure.  Yeah, I don’t…Yeah, using session, to me that’s a no brainer.  If you’re going to try to scale like this, if you’re going to go on any type of grid system, there are a handful of things you can do.  Session is one of them…

[10:53] Mike: Well, there’s a bunch of them.  I mean, you know, another thing that you’re going to have problems with is sending or receiving email.  So you have to use external services for that sort of thing, which means you can’t directly send an email, which means what you probably have to do is you have to cache it in a queue of some kind and then have that queue pulled.

[11:09] I’ve done a lot of parallel processing in the past, so that stuff kinda comes to me second nature, but I can see how if you’ve never done it or…

[11:19] Rob: Yeah.  No, that makes sense.  Yeah, and those weren’t the issues.  Those are things I would expect going into a grid system like this.  But the things he was talking about were actually like scale…they were config things and other things that seemed to me to be technical issues that need to be corrected.  You are never going to fix…you are never going to make session work or the email thing you just said.  I would never expect that to just work in a system like this.

[11:42] Mike: Well, I’ll let you know.

[11:44] Rob: Yeah, yeah.  I’m interested…

[11:45] Mike: I’ll do it in about probably three to four weeks, so at that point I’ll probably…over the next couple weeks I’m going to be experimenting a little bit more with Azure and seeing exactly what it can and can’t do.  And if I run into problems, I have no problems abandoning it.  I mean I’d rather not, but I haven’t done all the calculations on cost either to figure out what Azure is going to cost me.  I mean bare minimum I think it’s going to be $300 or $400 a month.

[12:11] Rob: Well, for a dedicated box you’re looking at least that.

[12:13] Mike: Well this would be for four.

[12:15] Rob: Right, OK.  Nice.

[12:17] Mike: So it’d be basically four.  The way Azure provides their SLA’s is you basically have to have a minimum of two for each of the different roles that you have.  So you have to have a web role, which is basically our front end web services, and then if you have worker roles, then you have to have at least two of them.  So that basically means you’ve got to have four and you’ve got to pay for four of them.

[12:36] I think they’re $90 a piece per month.  There might be smaller ones where you can get like shared versions of them which are $45.  So maybe it’s $180.  But still, I mean that’s not a bad starting point for release for my product, because my expectation is that my customers are going to be paying a couple hundred dollars per installation anyway.

[12:53] Rob: Right.  For the ability to scale that’s certainly kind of a no brainer.

[12:58] Mike : Right.  And plus, everything is all redundant.  So all the data that goes in is redundant.  All the queuing is redundant.  I don’t have to worry about anything failing.  Or at least presumably.  [laughs]

[13:08] [music]

[13:11] Rob: Well, now that we’ve bored any non-Microsoft developers to tears, shall we get into the nuts and bolts of this episode?

[13:19] Mike: Sure.

[13:20] Rob: It’s called The Customer Acquisition Process.

[13:23] Mike: So somebody asked me a few weeks ago, “How do I get customers for my products?”  It made me realize that not everyone knows or understands what the customer acquisition process really looks like.  So I thought we’d talk about the different phases of acquiring customers for a startup where your primary source of customers is via the web.

[13:40] Rob: Right.  And I think “via the web” is a good clarification there.  So obviously, there are some startups, if you’re selling stuff for $1,500 per license, your primary source of customers is probably not going to be via the web.  It might be, but it’s kinda tough to sell that with no-touch sales.  You get over $1,000, even if you get up into the…I mean we actually have quite a few people with Dot Net Invoice, which is about $300, who really want to talk to someone on the phone before they buy.

[14:08] So I think as so as you get up into the mid $500-$1,000 range and then up, you’re primary source is very likely not going to be the web, or at minimum is going to require quite a bit of touch via phone and potentially email as well.

[14:21] Mike: So one of the things to keep in mind is that not all of the products that you can envision are going to go through all these phases.  So any niche product is not going to go through all of these phases, because niche products don’t necessarily justify the expense of, for example, building a brand, which is one of the later stages.

[14:38] But phase one of customer acquisition, you can generally look at it as using some sort of AdWords strategy.  And this is a very early strategy because it can get you some initial customers, but it costs money.  And it’s not typically sustainable in the long term.  It can be, and six to eight years ago it probably was a sustainable strategy, but things have changed.

[15:00] As more competitors enter the space and start looking at the things that you’re doing, the feasibility of using Google AdWords or some of the other advertising models, it kinda falls apart because those people get into your space and they start advertising for the same keywords you are, which drives the cost up, and it’s very hard to keep the price to the customer down and compete on price if everybody’s driving all of your customer acquisition costs up.

[15:24] So one of the things you should do is definitely search the web.  Take advantage of any of the offers for free Google AdWords credit that are readily available.  You can get a lot of them through various email lists that Google tends to put people on.  If you sign up for some of their services you’ll get put on them.  Sometimes if you go to a web host, for example, they’ll have specials where when you sign up for their service, you’ll get a $50 or $100 in free Google AdWords credits, things like that.

[15:50] But phase one, again, it’s more of an advertising strategy.  Get out there, get initially in front of customers, get those first couple of customers.  And then from there you start figuring out what it is that they actually want and how to reach them on a regular basis.

[16:05] Rob: Yeah, I want to underscore what you said, because there are probably people out there listening that are like, “What?  Mike and Rob are recommending advertising?”  And that’s not exactly what we’re doing.  I typically don’t recommend advertising as an ongoing strategy, and that’s not what you’ve said.  You basically said use it because it’s quick to turn on and off, and you can get enough customers that you can really start iterating on your product and you can find out more about who your ideal customer is and isn’t.

[16:29] And I’ve mentioned this before, but I did a guest post on onstartups.com and it talks more about advertising and specific avenues.  Not just AdWords, but like Facebook ads and Reddit ads and StumbleUpon and things like that.

[16:44] But at the front of that I say exactly what you’ve said, which is use it to test.  Since you are able to turn it on and off quickly, you can spend $100, you can spend $500, and then you can just turn it off.  You can A/B test your pages, your marketing, your copy.  You can test pricing.  You can do all that kind of stuff.  You have a lot more flexibility.  And you can figure out which keywords are going to convert for you.

[17:07] Certainly, if you don’t have any money and you’re trying to launch something, you may not do this, but I have always thought that this is a valuable step in the process of honing your marketing message, essentially.

[17:18] And the next phase in this customer acquisition process, phase two, is SEO.  This is the strategy that is going to carry you for the long term.  It’s one of many strategies.  If you can get to the top of certain search engines, you’ve kind of made it.  It’s basically free traffic.  It does take time, effort, and solid analysis of your website, as well as of the people who buy, like what keywords they were coming through from Google AdWords.  You can then take those keywords and do search engine optimization.

[17:46] You really have to know what you’re doing to make this work.  It’s not something that ever really ends. Search engine optimization is absolutely an ongoing process.  But the idea is that you put in a bunch of effort and you get that snowball rolling, and then you put it into cruise control for a while.  And over many months you’ll watch your rankings slip a little bit, and then you can come back in and build a few more links and kinda get back up there.  If you neglect it too long, obviously you’ll have some serious ground to make up.

[18:14] But once you get to the first page and get towards the top of the first page for some decent traffic terms, it’s pretty incredible.  You can build a substantial business on this kind of traffic.

[18:23] Mike: Yeah, I mean the major difference between doing SEO and doing Google AdWords is that, obviously, on the surface, Google AdWords costs you money and SEO is free, but, to some extent, by doing Google AdWords, you find out the information you need in order to do effective SEO.  Because when you go into Google AdWords, there’s all these different tools that will help show you the keywords that cost the most and the ones that are searched on the most.

[18:47] And those will kind of direct your SEO a little bit, but it’s very possible that you also stumble upon a lot of the long-tail keywords which, in many cases, is where you’re going to get a significant chunk of your traffic because people aren’t necessarily advertising for those and you can end up at the top of the search results for those long-tail keywords, and it doesn’t cost you anything once you’ve figured out how to do that.

[19:12] But the key to that is definitely making sure that you think long term in terms of the SEO.  The AdWords are good for short term, SEO is for long term.

[19:21] Rob: Right.  And I would argue that SEO does cost you something.  I mean it’s going to cost you time or it’s going to cost you money.  It depends on how you do it.  But it’s going to cost you one or the other.

[19:30] Mike: Phase 3 is doing third party integrations.  And what essentially means is taking your products and integrating it with third party products, especially those customers whose customer base is substantially larger than your own.

[19:44] And that can be a great source of rather steady revenue.  Most of these major vendors have APIs that you can integrate into.  These include things like QuickBooks, FreshBooks, 37 Signals products such as Highrise and BaseCamp, and various others. And if you can add value for those customers, you are indirectly adding to the value of your own product.  And it gives you a lot of additional exposure on the websites of those vendors and, as a result, can be a source of additional traffic and a little bit more friendliness from the search engine.

[20:14] Rob: And the next phase is building brand awareness.  So Mike, you put this one on here.  Brand awareness is not so much about sales, but it’s making sure that when the customer thinks about the problem you are solving, they think of giving your product or company a try.  So it’s an indirect sales method that builds upon what you’ve already been successful at doing.  And if you haven’t mastered the previous phases, then this isn’t going to help much.

[20:37] Beyond that, is this like the social media type stuff?  Is that what you had in mind?

[20:41] Mike: Not really.  I mean if I were to throw out a problem domain, for example, let’s say you are in need of a database.  Can you name a couple of different databases?

[20:51] Rob: Right.  MySQL, SQL Server, Oracle.  Those are the three that come to mind.

[20:56] Mike: For any problem domain, you can certainly think of a couple of examples.  And you can usually only think of a couple.  That’s part of the issue.  When you get into databases, databases are so prevalent that that’s kind of a no-brainer.  It doesn’t necessarily meet the example very well.

[21:13] But if I were to ask you what sort of tools are out there that can help you manage or develop on SQL server?  Name off a couple of companies.

[21:21] Rob: RedGate, Microsoft.

[21:25] Mike: Well, Microsoft builds it. But aside from Microsoft; aside from the vendor themselves.

[21:29] Rob: I don’t know of any…I’m not a DBA, but gosh, RedGate is the only one I can think of.

[21:33] Mike: And that’s my point, is that because of brand awareness and exposure, you know who they are.  And there are a lot of others.  I mean, for example, Apex SQL is one of them.  And I’ve seen other tools out there, it’s just they don’t generally advertise as much, so I’m not necessarily as aware of their brands.  But they certainly make tools that are applicable to SQL Server.

[21:54] Rob: I see.  And this is why you put this at phase four, because this is something that this tends to be pretty hard to do and pretty expensive to do, I would think.  Because the only reason I know about Red Gate is because Neil Davidson has been putting on Business of Software for years.  Before that, I mean I had barely heard of them maybe four or five years ago.  \

[22:11] And like building brand awareness is like if you’re going to have a big…I can’t think of a small company that’s done this, right?   Big and small are relative, but to me, like I’m a small company.  Dot Net Invoice is small and we don’t have brand awareness, I wouldn’t think.  You know, people wouldn’t say, “What small business invoicing software is there written in .NET?”  You know, like no one thinks of DotNet Invoice.  You would go to the Google search.  We’re going to be one of these earlier phases that you mentioned.

[22:38] So is that why you put it here?  This is kinda maybe like once you’re big, more legitimate, maybe have more money to spend, have more time to invest.

[22:46] Mike: Yeah, it really is.  I mean it’s not something that you will generally do for the niche products.  And like I said upfront, not all products are going to go through all of these phases.  And any niche product is not going to go through this fourth phase because the market is simply not large enough to justify that type of a commitment.  Building brand awareness is more of an enterprise level product strategy.  It’s not for the smaller products that are only bringing in not even enough to support a fulltime income, for example.

[23:14] Rob: Right.  Interesting.  So I would say there’s another phase in here somewhere, or just another approach.  I brought it up earlier of like this social media marketing.  Honestly, think about App Sumo and how App Sumo has been marketed.  It hasn’t done Google AdWords.  It hasn’t done SEO.  It hasn’t done third-party integrations.  It has build brand awareness through like Twitter and through Noah knowing everyone and getting people to blog about him and doing that stuff.

[23:43] So maybe what App Sumo has done fits under brand awareness, but I think you said using social media is something else, actually.

[23:49] Mike: I think if you look at App Sumo specifically, there’s two others that I would say have definitely come across.  One is SEO and two is third-party integrations.  Third-party integration to me isn’t necessarily directly binding your product to those, but building your product and the marketing behind it.

[24:07] Rob: Yeah, that’s right.  It’s giving someone else a vested interest to link to you.  Yeah, I see.  So you’re right.  He does a lot of integrations.  I don’t think he does SEO…Well, I shouldn’t say that, but I’ve never found him through search engines.  I can’t imagine he gets much traffic on generic keywords.

[24:22] Mike: And that’s entirely possible.  But I would expect that the back links that he gets from various other companies probably count for quite a bit.  I mean let’s be honest.  If you’ve got a website that’s got incoming links from MailChimp and Twilio and Dropbox and all these other fairly popular products…

[24:42] Rob: That makes sense.  You are going to have a lot of clout in Google, so he could SEO if he wanted to.

[24:46] Mike: Right.

[24:47] Rob: So it’s interesting.  When I hear the customer acquisition process, I think of something different than this.  These are like…what are these?  These are like marketing approaches.  These are like getting people into the funnel, kinda getting them started.  So it’s a good summary of ways to get people into your funnel.

[25:04] [music]

[25:07] Rob: All right.  Looks like we have a listener question.

[25:19] Mike: We actually have two.

[25:11] Rob: Excellent.

[25:11] Mike: So the first question is about external markets.  This is from Egon.  He says, “Hi guys.  I really appreciate the realistic perspective that you guys approach in your podcast.  Excellent work.  I have a question for you.  I’m listening to Episode 26, so ignore this if you have a question answered in later episodes.

[25:29] I’m currently living in Brazil that has a very recent startup culture. I realize that in the US, most people start a company and have a huge upfront internet consuming market in the country itself.  That’s not always the case when you live in developing countries.  In Brazil, for example, it’s very difficult in some areas to have people pay for a SAAS app.”  That’s Software as a Service. “There’s a lot of cultural issues that I think are absent in other countries.

[25:53] For example, here a lot of piracy makes people undervalue software. You may think that this is my neighbor’s grass is always greener syndrome, but anyway, my question is, is it possible for a non-VC funded company to attack a foreign country’s market?  I feel that it could present a bit of a challenge, or am I wrong?  What could be the obstacles and what would you do?”

[26:12] Rob: So it sounds like Egon’s asking, when he says “a foreign country’s market”, he’s talking about trying to market something into the US.

[26:18] Mike: Yes.

[26:19] Rob: OK.  So this is common in websites that I see bought and sold, like on Flippa and elsewhere.  Traffic from the US is highly, highly valued because there’s so much money here and there are so many consumers.  So it is very common for people to start websites and apps from outside the US and market them to the US.

[26:42] So I think specifically, what Egon’s looking to do I think is perfectly acceptable, and I think it’s perfectly reasonable as long as…you know, he’s obviously able to write English well.  Obviously, language barrier could be an issue.  And as long as he’s willing to study up on the marketing and stuff, I don’t think there’s so many cultural differences that he couldn’t market an app into the US market.  What do you think?

[27:04] Mike: No, I totally agree with that.  The biggest thing that I would say that you have to be careful of is making sure that the language itself is not off by a little bit.  So one of the problems that most people have when they’re speaking any foreign language, even if they’re fairly fluent in it, is that when you get into writing, the grammar sometimes is just a little bit off.  And you can sometimes tell that it was written by a non-native speaker.

[27:31] And that can be off-putting to some customers.  If it’s very pervasive throughout the web-based application or throughout the downloadable application, if that happens to be it, it’s very off-putting to somebody who is trying to buy it.  But if you have somebody review that stuff who is a native speaker, you can easily get around 99% of those issues and it won’t be a problem.

[27:52] Rob: Yeah, you’re right.  Being 98% there isn’t enough, even if there’s a little weirdness in the marketing stuff.  It will put a lot of people off.

[28:00] Mike: Right, right.  So it’s good to have a native speaker who can look at that stuff.  But I totally agree with you.  I don’t see very many barriers to marketing a product to a foreign country.  Now, logistically, there could be some issues.

[28:14] So, for example, if you’re trying to take credit cards or you are doing online fulfillment, that sort of stuff can be a real problem.  And unless you are aware of the options, there is no shortage of companies out there that claim to have wide distribution patterns.  But you really have to start looking at what countries they service and which ones that they will be able to accept or not accept payments to and from.

[28:40] Rob: Yeah, I agree.  I was going to bring that up.  Credit card processing is such a pain.

[28:45] Mike: It’s a pain just living here in the US.

[28:47] Rob: It is.  And it’s way worse.

[28:48] Mike: I can’t imagine living in another country.

[28:49] Rob: That’s right, because all the services we have that make it easy for us don’t tend to accept stuff from people outside the country.

[28:57] Mike: Or they charge a lot more.

[28:58] Rob: Yeah, no, you’re right.  So I think when he says, “What could be the big obstacles?”, I think the potential language barrier and just getting payment and getting all that settled is going to be the big ones.

[29:09] Mike: Right.

[29:10] Rob: And, frankly, I guess hosting.  Like if you host in Brazil, I would probably not recommend that, I think for two reasons.  One, because there’s going to be latency.  And if everyone is coming from the US, when I go to sites that are outside the US, they are a lot slower for me.  So I would probably say host in the US if you can.

[29:25] [music]

[29:28] Mike: So the next question is from Ruford.  He says, “Hi guys.  I’ve been going through Rob’s book and I have a question regarding the market research part.  I live in Costa Rica and I am thinking of making a hotel management software product for family owned or small hotels here in Costa Rica that offer specific benefits to Costa Rican benefits.

The Google search volume for Costa Rica for related terms is really small.  However, I believe that this is because the products I want to build is not something thousands of people would look for, but is really very specific for small hotel owners.

[29:58] There’s at least one company that provides a similar service, and according to their website, they serve about 20 hotels.  The market of all small hotels in Costa Rica has to be above 200.  Given that the Google search volume data is not really useful, can the existence of this company and the small hotel company enough for me to say that there is a good enough market for me to start developing this software?  Thanks a lot, Rupert.”

[30:19] Rob: OK, so we need to do a little bit of math here.  I’m wondering how much he’s going to be able to charge.  Let’s say he’s able to charge $20 a month for this software.

[30:30] Mike: For hotel management?  You think that low?

[30:33] Rob: I would guess, yeah.  If all it is is booking and it’s an online service, I’ve seen this for small hotels.  I saw one for sale on Flippa and it had like a $15 plan and a $29 plan.  I think they may have had a $39 plan but it didn’t go up much beyond that.  I mean these are Mom ‘n Pop things.  These are hotels with…they’re kind of like B&B’s; that’s what I’m thinking of—small inns with like half dozen or a dozen rooms.

[30:59] Mike: OK.

[30:59] Rob: Yeah.  So it’s almost the consumer market.  It’s like one step above the consumer market.  So I’m not saying that he can’t charge more than that, but my guess is that especially in Costa Rica, where it is cheaper, just cheaper to live, your US currency will go farther.  I think you have a tough time charging $40 a month down there.

[31:14] My guess is…let’s just throw out a number and say $20 a month.  If the existing people only have 20 customers, you’re talking $400 a month that you can make from this.  And if you get every hotel in Costa Rica, or every small hotel, you’re only talking four grand a month.

[31:29] I don’t know.  I guess you have to answer the question.  It’s like can you build it for cheap enough? Or can you build it fast enough that amount of money makes this worth your while?  I think that the fact that the first provider only has 20 customers makes me think that these customers are probably pretty hard to reach.  You know, are you willing to do cold calls or drive out and meet with them and kind of get them onto your system?  My guess is they’re not super technical, and so you’re not going to be able to sit at home and put up a Google ad and have people come through a funnel and convert.  I think it’s going to be much more of a custom sales process.

[32:01] That’s not to say that that’s a really bad thing. It’s just know that upfront, that you’re going to have to leave your house, very likely, and you’re probably going to have to either talk on the phone or meet face to face with every customer you sign up.  That would be my guess.  Because since no one’s searching on Google for it, I just don’t know how else you’re gong to get traffic that that’s specific.

[32:18.7] Maybe you have a list and you could do mailings to all of them.  There are other options here, but that’s my take on it.  What do you think Mike?

[32:25] Mike: I think I would go to that potential competitor and see if you can find a couple of the hotels that they actually service and then go talk to them and find out how much it costs them.

[32:35] Rob: Right.  Call them up and act like a potential customer.  Say you have an inn…

[32:40] Mike: No, I was saying, you know, contact that other vendor’s customers and ask them questions.  Say, “What is it that you’re actually using this for?”  Or, you know, if you can identify a couple of the customers, then you can start looking at them and saying, “OK, how many rooms do you have, and how much are you charging for those rooms, and what is your approximate revenue per month?”  I mean you can ballpark a lot of these numbers just based on room counts and how much they charge.”

[33:03] And those things, I think, will give you a little bit better idea of the actual market and what market your competitors is serving.  Because until you know that, I think that you’re kinda shooting in the dark.

[33:16] Rob: That’s a really good point.  I do think it’s all about doing the math upfront and figuring out whether it’s going to be worth it to you.  Because you’re right.  If they are able to charge $99 a month, then that becomes lucrative pretty quickly; lucrative for a micropreneur.

[33:31] Mike: I mean yeah, if you can charge $100 a month, if you’re serving 20 hotels, that’s…

[33:36] Rob: Two grand.  But two grand, I mean you only have to sign up 20 hotels and that’s recurring revenue.  I bet your lifetime value of a customer is going to be really high, because once you get them signed up, and if you provide good service, they’re not going to have much motivation to leave.  So I bet they’ll stick with you for several years.

[33:53] So it’s probably worth the effort to actually meet face to face or talk over the phone if you have a several thousand dollar lifetime value for that customer.

[34:02] Mike: I mean the other thing is you can also look at some sort of a model where you get paid based on the bookings that you do if you do a little bit of marketing or advertising or you take the reservations on behalf of that company.  Kinda inserting yourself as the middleman, that’s another option.  Maybe it is, maybe it isn’t.  I’ve never actually worked in this particular industry before.  But there’s other ways of attacking the problem and trying to monetize it as well, as opposed to charging them directly.  Maybe you only charge them in certain instances.

[34:31] Rob: I think the other thing to keep in mind is that almost all of these owners are not going to be technical people.  And so the support burden is going to be…I won’t say substantial, but there’s going to be a lot of questions.  You know, basic stuff like, “I click on this and it doesn’t work.”  It’s going to be some pretty basic non-techie questions.

[34:51] Again, that’s not to say you shouldn’t…

[34:54] Mike: Non-techie descriptions.  [laughs] “This doesn’t work!”

[34:55] Rob: Yeah, exactly, when it’s like, “Well what does that mean?”  Again, not saying you should not do it, but realize that the support burden for these types of customers is going to be a lot higher than if you were catering to web developers or designers or something.

[35:05] [music]

[35:08] Mike: If you have a question or comment, you can call it into our voicemail number at 1-888-801-9690.  How many more weeks do we have until we decide to scrap that?

[35:18] Rob: The podcast just came out today where we basically said, “Call soon or we’re going to take our number off.”

[35:26] Mike: Or you can email those questions in MP3 or text format to questions@startupsfortherestofus.com.  If you enjoyed this podcast, please consider writing a review in iTunes by searching for “Startups”.

[35:35] You can subscribe to this podcast in iTunes or via RSS at Startupsfortherestofus.com.  Our theme music is an excerpt from “We’re Outta Control” by MoOt, used under Creative Commons.  A full transcript of this podcast is available at our website: Startupsfortherestofus.com.  Thanks for listening and we’ll see you next time.

Twitter Digg Delicious Stumbleupon Technorati Facebook Email

2 Responses to “Episode 43 | The Customer Acquisition Process”

  1. Mike, regarding to yours SQL datetime2 issue. You may try to change a compatibility level on your local database (Database –> Properties -> Options -> Compatibility level) and set it to SQL Server 2005. This should make your database behave as it is hosted by 2005 server.

    Great podcast guys! Keep going!

  2. Oleg, that’s a great tip and thanks for the suggestion. I’d found that, but it’s not a database that I have direct authority over. It would have taken a change control procedure to modify and the easiest way to handle it (for the time being) was to change the ProviderManifestToken instead.

    At the time of the podcast, I didn’t have time to make the attempt, but we’re in the process of working through that change control procedure. If it works, then I shouldn’t have to worry about importing anymore objects into my EF model and wondering whether it’s going to break something or not.