In one of our updates we realized we introduced a bug that caused a user’s scanner to stop working after adding or editing a picture. We had heard reports from users that their scanners weren’t working but we couldn’t figure out why. If we had them reset their app in the system tray everything worked fine. Finally, we figured out how it was happening.
We have a fix submitted to the market and it should be accepted by Apple within two or three weeks. So don’t update a product photo unless you know how to reset an app in the system tray until you download the update (just double tap the home button and delete the app from the system tray as seen to the right).
Obviously this sort of bug is HORRIBLE because it impacts our most dedicated users – i.e. those who edit and update products, retailers and prices.
Thought I would share this picture of John representing ShopSavvy at the Shopper Marketing Summit in Chicago. Great job!
Several folks emailed me asking why we didn’t go down because they know we use Amazon’s Cloud for ShopSavvy. Last night Amazon’s cloud crashed taking down a number of other companies. Great news, ShopSavvy did’t go down. Two weeks ago I wrote a post detailing our server environment locally and in the Amazon Cloud. The easy answer is that we have two sets of infrastructure hosting our backend – if Amazon goes down our data center takes over and vice versa.
This morning I woke up to find that someone had commissioned a comparison between ShopSavvy and ScanLife. The post was titled “UPC Scanner App Comparison: ShopSavvy vs. ScanLife” and was found on the From One Mommy to Another. From what we can tell the two important guidlines of the test were:
- Comparison to take place on the HTC Thunderbolt
- Shopping to take place in Nordstroms
When I read these two guidelines I knew we were in trouble. First, HTC’s Thunderbolt has a big problem with garbage collection causing apps that use large pictures (like ShopSavvy and Facebook) to crash. While HTC is working on a fix, in the meantime Thunderbolt owners are really struggling to use ShopSavvy. Apps like Scanlife that don’t use large pictures or heavy graphics don’t crash as often. Second, Nordstroms provides us a feed of their pricing and inventory data, BUT their products are often unique to their stores. Many of their items can ONLY be found in Nordstroms making a PRICE COMPARISON app basically useless since other retailers don’t sell the other products. Based on these two factors I assumed we were set up to fail.
But when I scrolled down to the bottom of the page (and let me tell you, it was a long article) I was shocked to learn that the author actually picked ShopSavvy – HANDS DOWN! Here was her quote: “Overall, ShopSavvy wins this comparison hands down. It is as easy to use as ScanLife, but the organization and added features take it beyond being just a basic scanner.”
I was floored! I can’t wait until she gets the HTC fix and goes to a store with products sold by multiple retailers.
When Matt Weathers joined the team he wanted to know how well we were performing in various areas. He has hired a number of third party firms to research how consumers use ShopSavvy, how they use competitive apps and how well we perform on local inventory and price search. The data coming in has resulted in big changes in the app itself. For example, we simplified the UI when we the research revealed users were completely confused by the old UI.
The local research part of the study was conducted by a former Deloitte Consultant and is about 50% complete. Our plan is to release the results in a white paper, but I thought I would share some of the early numbers coming out of the research. We think we are doing really well revealing local inventory and pricing information to consumers and the research is bearing that out (hell, we’ve be doing this full-time for three years). In the graphs below you will see we focused on the number of retailers we returned during the study as well as the price returned
One of the hottest new phones from HTC, the Thunderbolt won’t run a LOT of apps like ShopSavvy. We wrote about the issues extensively last week here. We reported our findings to Google and they have confirmed it is an HTC issue. We have not heard back from HTC, but it is clear LOTS of developers and users have let them know about the problem. The Droid had a similar issue when it was launched and it took Motorola about three weeks to issue a fix – we are hopeful HTC will be able to resolve the issue shortly. In the meantime, ShopSavvy won’t work. We are updating the app to send a message to Thunderbolt users explaining the issue and letting them know we will alert them when HTC fixes it. If you are Thunderbolt users PLEASE stop giving us 1 star ratings in the various app stores – we know it doesn’t work and there isn’t a way for us to prevent you from downloading the app.
Apple and Google just updated their ranking systems in their respective app stores and based on what we are seeing they both decided to punish apps that ‘game’ the system. Now the stores seem to check to see how often your app is downloaded AND how often it is used by those who download the app. So if you are paying folks to download your app you will get punished HARD. As a result ShopSavvy’s rank in every market has improved and our daily downloads have almost doubled. Here is a run down:
- #6 in Apple App Store (Utilities)
- #5 in Android Market (Shopping)
- #3 in Amazon’s Android Market (Shopping)
- #1 in Windows Phone 7 (Tools)
UPDATE: GigaOm is reporting that Apple is now rejecting apps that use Pay-per-install campaigns.
It seems like every third comment we get on ShopSavvy in the Android Market is spam from ShopKick. Here is an example:
The first example from MaryAnn. She gives us a 5 star rating and then gives us a compliment, but then suggests downloading ShopKick using a code included in the comment. The second example is from Kathleen. I don’t like Kathleen as much because she rates us 3 stars. Most of the ShopKick ratings are 3 stars.
You may ask, “Why is ShopKick doing this?” The answer is simple. The app promotion companies they hire to drive downloads target popular apps like ShopSavvy and insert promo comments just like these. Do they work? I have no idea, but ShopKick is spending millions to keep their app highly ranked.
The best part of the story is that Google is wise to ShopKick. The Android Market’s comment spam system does an AMAZING job of filtering these out. It takes a day or two, but Google eventually finds most of these and deletes them (hopefully they delete the ratings too). Great job Google!
You know how some people Google themselves? Similarly app developers like us spend a lot of time checking our rank in various app stores. For example, ShopSavvy has dominated the Android Market, but we have been runner up to Red Laser in the Apple AppStore. We are please to note that is changing (organically, no paid downloads). As of this evening we are more popular than RedLaser in the Apple AppStore! In the Android Market we are beating Google Shopper, ShopKick, Groupon, ScanLife and Amazon. In Amazon’s market we are beating both Barcode Scanner (Google) and Ebay – one spot from beating Amazon in their own store! In the Windows Phone 7 market we are killing everyone in our space. Rock!
If you are on iOS take a minute to update and rate us again, here is the link.
For some reason we never got business cards printed for ShopSavvy. I think the main reason is that we don’t have a logo. We came up with an icon for the app (we are on number three since launch), but no actual logo. Showing up at a meeting without a business card sort of sucks, but who uses business cards anymore? With the web it is really easy to track down contact information, right? Well Matt and John really want cards so I had these printed:
The front of the card has a scanable barcode that links to a product associated with the owner of the card. The back of the card has a QR code that contains his contact information. You may notice we STILL don’t have a logo. Does every startup really need a logo? Craigslist doesn’t have one. Google really doesn’t really have much more than a type-treatment logo. What do you think? No logo, no big deal?
