Take a look at how other companies with similar problem resolved it, for example: AirBnB, Uber, TaskRabbit and others. Every one of them, unless they were extremely lucky with timing, used a narrow geographic focus, approaching established communities and fulfilling demand for one side by hand for the first little while. In your case you can approach established organisations like universities, churches, foundations, volunteer organisations and the like to get people that would do things and do exactly the same for the other side. Initially you need to do it by hand until it starts going viral. Spreading the word through social media is great but it needs to be very targeted and approach people that can already find something in the application. If there are other organizations that do similar postings, for example if there are requests for volunteering on Craigslist, you can approach them to have it posted in your application. Narrowing on one geographic location is always easier in the beginning, because you can learn the area and all the players involved in the scene, and then maximise the output, expanding to new geographic locations when needed. It doesn't mean limiting your app, but just focusing your marketing efforts in one narrow location that can bring biggest output. Everything that I mentioned above can be applied to offline and online social marketing. Unfortunately there is no magic solution to your problem except for bootstrapping it in the beginning and finding communities that you can leverage with biggest social impact return. Good luck! Denis.

The best way to approach this problem is by identifying primary pain points that users share and focus on those feature-wise, a common denominator of some sorts. These pain points need to be very high/strong though to be valuable for multiple target audiences. If you're facing two target markets (you're a connecting middleman of some sorts) then you might consider building two separate entry points with two functionality sets. That is only justifiable when one target market offer cannot exist without the other. In a typical situation you still want to identify the target users that contribute to the biggest market with largest growth potential and tailor your offering to those. Once you identify the segment try to understand who early adaptors are and make your product for their biggest pain points first, then expand once you have traction. As a conclusion I would suggest that sometimes you have to choose who you leave behind in your offering at least until you have traction and growth with your primary target audience. Please feel free to reach out if you want more advice, as I deal with that on the daily basis with my current job. Cheers, Denis.

Back in 2010 we went through exercise of first using a local data center and then using a CDN in a different country. Interestingly enough using a remote CDN had much smaller latency. After doing an in depth analysis this is what we discovered. There are two factors that influence your latency: throughput of the pipe and speed of the physical server. Let's start with throughput. Consider two scenarios. In the first one you're using CloudFront from Amazon (or any other big name CDN) which probably has direct connection (or near direct) to the continents fiberoptic pipe. The second scenario is if you build a data center in the country. In this case you will be removed from the main pipe by several hops through internet providers which in turn will narrow the throughput after every hop. Unless you are willing to pay very expensive bill for the access to continents, or at least contries main pipe the latency will be higher than from big name CDN's. The second aspect deals with the type of servers involved in servicing GET requests. You need to install top notch hardware with very speedy SSD hard drives and extremely fast networking cards. You can go as far as start optimizing bus speeds between the components. All that leads to increase in price per server. If you're using a CDN then you're much better off because all of that is taken care of for you and you constantly run on latest and greatest hardware that you don't need to upgrade. Count in the amount of time/money need to be spend to service a data center. From my experience the companies that do administration on demand are slow for near 100% uptime requirements so you will need to hire and staff to do monitoring and maintenance. Also never forget about redundancy that also needs to be provided as your server will fail and will go down. Security is also a concern. My conclusion from the the evaluation done in 2010, which is much stronger today, go with CDN unless you're a Fortune-500 company and need a dedicated special use data center. Current CDN's usially offer multiple data centers at different continents for your disposal that can further shrink your response time. Good luck! Denis.