The Snowtide Blog

**Marketing is really hard, despite the rumors you’ve heard. The more I get into it, the more I’ve come to respect the skills (if not necessarily the tactics) necessary to deliver a message to prospective customers.**

Up until this point, Snowtide has done virtually no marketing, and we’ve made out very nicely. We now have a mature product that really kicks ass. I’m proud of what PDFTextStream is doing for its users, some of whom simply would not be able to do their jobs if it weren’t for it.

But we’re past the point of working small niches. Scores of development shops, large and small, would have fewer bad days if they had PDFTextStream humming on their servers and in their products. So, the time has come to spread the gospel and make sure they know that.

To that end, we’re starting a new marketing strategy in July. It’s going to start slow as we learn our footing (the conventional wisdom is that summertime sees a slowdown in corporate software purchases because of vacationing). It will build through the end of the year. And, it will end with PDFTextStream being the only serious choice for developers in enterprise-class environments.

There’s the tricky part, though: convincing people that our product is better than its competition. The foundation for that has been laid for PDFTextStream — it’s been borne out in customer experiences. The problem is that, without appropriate marketing, the people that are likely to appreciate that fact will never even know about your product. In order to change that, we’ve got to write good ad copy, hire good designers to craft and mold that copy into digestable elements (ad banners, text ads, white papers, editorial placements, etc), and feed those elements into a cacaphony of interruptive marketing noise to be noticed and not ignored.

Technical people and marketing folks have always had their differences; they simply do not understand the difficulties inherent in their respective trades, and that often leads to disrespect. That is ever so slowly changing, in part because of pieces similar to this post, typically made by an in-the-trenches software company founder (like myself, I suppose), who inevitably describes how difficult marketing is. And seriously — it’s really, really, hard.

Every step in the progression of tasks I enumerated that leads to a prospective customer seeing, noticing, and acting on a pice of advertising is hard. And personally, I find it very unpleasant, simply because I am, by nature, technical. I know how the bits in software work, and I know those types of things very well. It’s a perfect occupation for someone who is a bit of a control nut. Yes, I am that.

So it makes me very uneasy to engage in an activity (like marketing) where I cannot readily control the outcome. It makes me even more uneasy to engage in an activity (like marketing) where I am less than fully confident in my (and in this case, our) abilities. We are fundamentally technical; we know how the bits work. Even with help, we find the fuzzy, soft, vague world of marketing just a little scary.

That will get better in time, as we fail a little, succeed a little, and do a little more of the latter and a little less of the former each time we try. It would be a high crime to not try, try hard, and try often; we have a great product, it should be seen, and it will be seen.

Many times I’ve been told, ‘Snowtide should forget about doing custom development work, and concentrate on selling product’. Of course, few people realize just how important our custom development clients are to our overall business and to the quality of our product.

It should all be very simple, right? We have a great product (PDFTextStream) that does a fantastic job of extracting text and metadata out of PDF documents. It’s an ideal solution for Java developers that need to integrate PDF workflows into their desktop or web applications. Many of the more technically-oriented circles we travel in think that that’s enough, a clear indication that they fall prey to the myth of ‘if you build it, they will come’.

However, one of the ongoing keys to our business is how we maintain a divide between customers and clients:

• Customers keep the trains running on time. They purchase licenses for PDFTextStream as-is, and require very little hand-holding and advice — after all, they’re software developers themselves, and in general know what they’re doing.
• Clients form our inspiration. They need the base functionality of PDFTextStream, but it doesn’t do anything for their business needs alone — our clients by necessity look to us to help them build features and application functionality on top of PDFTextStream that will give them a competitive edge in their industry.

It’s no secret that there’s a lot more money to be had in selling product to customers instead of building new applications for clients. However, few understand just how important the latter are in informing our product development strategy, which makes PDFTextStream that much more attractive to new customers in more and more industries.

As it stands, PDFTextStream incorporates a number of features that have been specifically requested by various clients from various industries. And, of course, we’ll continue rolling out new sets of features that complement specialized needs. So, we thank our clients; without them, we would never have made it this far.

**As much as we love getting positive feedback, we like constructive negative feedback even more. So, when someone let us know that PDFTextStream didn’t meet their expectations, we moved fast to change that.**

When we released PDFTextStream, we were pretty confident in its abilities and its completeness. There’s a lot of *stuff* to handle when building something that is going to be consuming and processing PDF files — there’s some significant variability in how the PDF specification is implemented by various document producers. We obviously want PDFTextStream to handle anything that is thrown at it, and I think we’ve succeeded to a large degree (at least, the excellent feedback we’ve received so far is a very good indication).

So, you can imagine our reaction when we received an email from Leonard Rosenthol about a week or two ago. Leonard is the CTO of [PDF Sages](http://www.pdfsages.com), a very well-known consultancy that specializes in PDF document products, services, and workflows. It turns out that Leonard, after running some PDF files through our online demo, was *disappointed*.

As much as we love getting positive feedback, we like constructive negative feedback even more. Leonard was disappointed because PDFTextStream didn’t support XMP ([Extensible Metadata Platform](http://www.adobe.com/products/xmp/main.html), Adobe’s XML-based embedded metadata tagging system). Now, we knew this, but we were planning on implementing support for it in some later version of PDFTextStream since XMP metadata streams usually just duplicate the metadata that is otherwise available via other sources in PDF files. For the most part, extensive use of XMP metadata streams is confined to custom workflows, higher-end prepress environments, and other areas that we were not planning on addressing for some time.

However, all that doesn’t make Leonard less disappointed, and I knew that it would be completely irresponsible to rationalize his concerns away. So, we did just the opposite: we scrambled, worked our tails off, and added XMP stream extraction support to PDFTextStream (starting with yesterday’s release of v1.2).

So, what’s the point? We care about what our customers think and what they need. When one of you is disappointed, we want to hear about it, and fix the problem, ASAP. Sure, that’s probably clichéd notion, but we actually try to live up to that ideal. I hope someone notices. . .