Hi guys,

I'm working on a site (http://www.soapservice.co.uk/) which is pre launch at the moment, but which is fully functional (except registration). You can go see it - it has a lot of useful SOAP data stuff, like ip2country services, zip code services, distance between two zip codes or post codes, currency conversion, that kind of thing.

What I'd like is, maybe some of you could go check out the site, see if you think you'd use it... I'm also looking to give out a bunch of free passes, maybe for a few months of use. I have a bunch of servers in datacenters now ready to power this, but if you guys would like to super power your sites for free and give it a test, that could be mutually beneficial.

My immediate thoughts:

1. Those services have nothing to do with SOAP. May want to be careful how you market yourself (may put off developers if you sound like you don't know what you're selling).

2. Most are available for free through other services. You should focus on your value proposition: a) we make it easy, or b) guaranteed uptime, or c) something else that makes your service seem worth 5 quid a month.

Steve,

I don't think any services are to do with SOAP itself, that would be a tad dull. All the services are available via SOAP however, with clear examples given for every single one.

Some of these are available elsewhere; not really the card to bank lookup thing. And none all in the same place. And none with a quality of service aspect. Would you really launch a site and have a lot of different free sites propping up your baby?

I reckon it's worth a lot more than a fiver a month - try hosting the same yourself and watch your server fees stack up. But we can always agree to differ. The market will ultimately decide.

Jsecure,

I'm not saying it's not worth it. I'm saying you have to pitch the value proposition if you want to sell it. Hope you understand the difference.

PS - Post your link at www.cloudomatic.com to advertise it to potential customers.

Steve,

Sorry if I sounded a little guarded. Forums can tend to be negative places. I think you're right - the ease of use and guranteed uptime will be key selling points. I think the third, and possibly the biggest one, is that this kind of business data will simply be impossible for some people, either because they lack the technical know how to put it into database form and get it going on a server, or they lack the finance... But there are a lot of talented designers who could use access to this, but simply it's not an option for them.

My site should give them a business oriented provision of that data. Not a 'nice free service', but a guranteed business service. We guarantee to be so good, you can run a business from us.

Thanks for the link - that's actually a really useful place to pitch my site. I wasn't aware of that until now.

No worries, I know a lot is lost in translation in text.

Another suggestion: If you're targeting designers or novice developers, you should also create a client plug-in for PHP (and possibly for other languages) to make it simpler to use your service (e.g. from a simple function call).

£50.00 /month
30,000 lookups / month

If thats IP / Geo location lookups...?

My server / site does that amount every 5 hours or so..

Maybe 30k is a bit low? dunno.

Still....
The site looks very nice and i wish you the best.

Sorry to pick holes...

But your prices also dont add up?

You mention:
Our rates are as low as £0.001 per service request, which means a thousand requests would cost just £1.

However your selling:
£5.00 / month
500 lookups / month


Surely that would be 5,000 /pm Or slightly more as its a subscription?




Just thought id point it out.

lynx, if you're doing 6k geo lookups per hour, you may need to revise your code. I'll usually store geo ip in a session variable for further use to prevent a lookup on every page view. Typically results in a 90% reduction in lookups.

Originally Posted by smoseley

lynx, if you're doing 6k geo lookups per hour, you may need to revise your code. I'll usually store geo ip in a session variable for further use to prevent a lookup on every page view. Typically results in a 90% reduction in lookups.

The lookups are to a mysql DB that has the Inetnums in the table.

the 6k/hour is because the server gets ~6k /hr unique hits.

Lynxus,

The prices don't add up purely cause it's pre-release and the final trawl over the site hasn't been done yet. The headline prices are pretty much what I'm thinking of going for... You'll notice there are no Ts & Cs yet and other things. That all needs to be done. Functionally though it's getting close - you can do a bunch of lookups already.

If you have a site that's getting 6,000 hits an hour, it's probable you're not the target audience I'm going for. If you're that big, having your own geolocation database must be easy (there has to be money behind 6,000 hits an hour, or the site is just wrong, wrong, wrong, somehow) and sending all that traffic to a third party to process may simply just not be a smart way to handle things.

I'm not sure there's going to be a really price efficient way of handling that much traffic to be honest, without giving dedicated servers for the job. Long before you get to that point, it's more efficient for a company to hire a contractor, or their own staff, get a CSV of geolocation data, and just do the job themselves...

My site is not to argue people shouldn't do it themselves; it is undoubtably the cheaper option if you are big... My site is to say, for some people they don't have the need to shell out for the whole infrastructure of doing it themself, and it would be nice if for say, £30/month, someone could give them some of the really juicy bits of business data (zip code distance comparison, currency conversion, postcode lookups, credit card lookups...) without them having to do anything or run anything or maintain anything...

Hi there,

I just tried out your postcode distance checker as I would have actually made use of that, however, it doesn't work correctly - example - http://info.soapservice.co.uk/servic...QZ/to/BT126LA/

The distance between these 2 points is below 3 miles, however in this instance it's returning as over 30.

Thanks,
Andy

Andy, are you sure that it's 3 miles from center to center and not perimeter to perimeter? Also, are you sure that's returning miles and not km?

Regardless of what it's returning, it's entirely incorrect.

http://maps.google.com/maps?f=d&sour...=UTF8&t=h&z=14

Looks interesting, very handy for webmasters with limited data sources.

Originally Posted by Andy Pugh

Regardless of what it's returning, it's entirely incorrect.

http://maps.google.com/maps?f=d&sour...=UTF8&t=h&z=14

It's not incorrect.

The database we have makes use of postcode outcodes. That's stuff like the E15 in E15 XXX. It's likely Google is using the full postcodes, which allows for a slightly more fine grained sense of distance.

As it happens, I have just bought a postcode directory (FULL postcode directory) for my site. That moves it from using a 2,800 record outcode directory, to a 1,600,000 record full postcode directory.

It should be interesting to run your test again once that is live. I'll let you know.

I'm not sure that the full 1.6m record database will be used in the final thing though. Where outcodes fail on tiny distances because they are big shapes instead of tiny ones, measure a distance of substance, say 20, 30 miles etc, and the tiny 2,800 record outcode database will perform almost the same as a much bigger, potentially slower 1.6m record one.

On the other hand, maybe I'll decide the 1.6m records are worth it. I'm going to see how much difference they really make. If the new record set excels in your test, I might keep it in. If people are paying, why not provide a top notch service, I guess.

Thanks for posting your test, it's interesting. Data is rarely wrong, it's just different. And more often than not, there's a sensible reason why.

Hi,

No problem, to be honest with the 2800 record database I couldn't use that on a live environment where I strongly rely on solid data, for example - if it's only taking the first 3 digits that's the difference between BT23 and BT2 - both 25 miles from each other.

Thanks,
Andy

Andy,

Can I ask, what is the business you need high precision location data for? I want to understand a number of different clients and a number of different useful implementations of my data when the site officially launches, to kind of pre-empt what people might want to do with the data, and design services to fit some common use cases as much as possible.

If you could tell me what you use the data for, that would be helpful. I can imagine such things as finding the nearest store to the customer, or the distance between two people... Those wouldn't really require high precision data, as to give a general distance '20 miles' etc, would be alright (they're not going to measure or anything...) but perhaps you have a reason why high precision data would be needed.

I'm genuinely interested; I'd be pleased if there's another good reason to use my data I can add to what I have already...

Originally Posted by Andy Pugh

Hi there,

I just tried out your postcode distance checker as I would have actually made use of that, however, it doesn't work correctly - example - http://info.soapservice.co.uk/servic...QZ/to/BT126LA/

The distance between these 2 points is below 3 miles, however in this instance it's returning as over 30.

Thanks,
Andy

Andy,

The new 1.6m record postcode database is up, and it seems to run pretty fast. I can't notice a difference between that one and the old one. What I have is a hybrid solution; it first uses the 1.6m database, but falls back to outcodes if what you entered is not a recognised full postcode.

Anyway, the two codes you describe now resolve as 1.74 miles apart, probably much more like what you're looking for.

So listen, your test was quite helpful, and you mentioned you might use my service for a product of yours. It would be a help if I had somene making use of it - could I offer you a free £30/month account for three months? Would you use that?

deleted