AirPods - One Year On

AirPods, Apple's take on truly wireless earphones, have been out in the world for just over a year now. They were plagued with 4-6 week wait times right from the initial launch. These shipping delays continued throughout the year, and are now affecting 2017 Christmas holiday shoppers. Whatever way you look at it Apple just isn't able to make enough to meet demand. Whether this is because they're difficult to manufacture, or because demand is higher than expected, no one is sure. One thing is certain, however. AirPods are popular. You can't walk a block through the Sydney CBD without seeing someone wearing a pair, and it isn't uncommon for me to not be the only one on the bus wearing them. Alongside the Apple Watch Series 3, AirPods are my favourite Apple product released in years.

Now, one year on, I want to look back on what I wrote in my AirPods review and explore how my initial impressions have either changed or remained the same.

These are more than special earphones. They’re a new category of wearable device.

I still believe that AirPods have the potential to be considered a wearable device, and not just a pair of earphones. Over the last year thanks to software updates, transitioning the connection of AirPods between devices (iPhone, to Apple Watch, to Mac, to Apple TV) is more seamless than ever. If I'm wearing them, it doesn't matter if I press play on audio on my phone or watch, the AirPods intelligently will connect to whatever device is best. This has other implications, such as always being able to communicate with Siri. While wearing AirPods, I know I can activate Siri discreetly on my Apple Watch, and talk as though I was holding my phone near my face.

They haven’t once felt like they’re going to fall out and this includes while at the gym doing all of my regular exercises including running, weights, pull-ups, push-ups, sit-ups; you name it. They’re hardly noticeable in the ear.

It's easy to forget you're wearing AirPods. They remain as lightweight and comfortable as ever (though the comfort factor isn't the same for everyone), and despite not having perfect sound quality, the convenience of the AirPods design is enough to make me not want to leave the house with any other pair of earphones.

Although battery life isn’t fantastic for “wireless” earphones, my early usage suggests 5-6 hours on a single charge - in line with Apple’s estimates.

After using AirPods for a year, I'm convinced that the battery life does not matter. I'm not sure how long they last on a single charge anymore; it could be 2 hours, or it could be 7. It doesn't matter. There is almost never a situation in which I go more than 90 mins between putting them back in the case to charge. I've never found the battery life not to be good enough. Sure, a few hours may not sound like a lot, but how frequently are you using them continuously for that length of time? Even popping them back in the case while you visit the bathroom at work is enough to top them up for a few more hours of use.

Having to charge so frequently might sound in conflict with what I said earlier about wearing them constantly.

Still true. While battery life is good enough for day-to-day audio listening, the idea of AirPods as wearable tech won't be taken seriously as such until they can last in your ears all day should you need them to.

Siri on AirPods has been a disappointment for me so far. The way to active it is to double tap on either bud and then talk.

I've found the best way to talk to Siri with AirPods in is to activate it from the Apple Watch by pressing the Digital Crown. It's going to work every time.

To truly be free from your phone while using AirPods, a smartwatch with audio controls of some kind is necessary.

This is true more than ever. Controlling the AirPods with only a phone is still clunky, but using a smartwatch makes this a lot easier. Now that the Apple Watch Series 3 has a cellular/4G connection, the combination of AirPods + Apple Watch is even better. You can now leave the house without a phone and listen to music while walking around or exercising, while also being able to take phone calls, and dictate messages. It's a powerful combination, and having the AirPods with the cellular watch makes the experience notably better.

We've seen some small improvements in the last year to both the software controlling AirPods, and to the hardware they connect to. Improvements to the way AirPods handle connecting to different devices and new hardware such as the Apple Watch Series 3 take away some of the hassles I addressed in last year's review. There are a few improvements I'd like to see to AirPods going forward. The first is better sound quality. The sound quality of AirPods isn't bad by any means, but there are times when higher quality audio is appreciated, and not having to find another pair of earphones in these situations would be nice. Secondly, water resistance. iPhone and Apple Watch are both water resistant enough that if you're stuck in the rain with them it isn't an issue. It would be nice if AirPods were also water resistant. I've heard that they do survive submersion in water, but until Apple promote water resistance as a feature themselves, I'm reluctant to let the AirPods get wet. Finally, I'd like to see a dark (space?) grey or black colour offering - purely for cosmetic reasons. I know that Apple are unlikely to stray from their distinctive white earbud design, but the new Space Grey peripherals that come with iMac Pro show that they are open to mixing things up.

CGM Diaries: Week 1

I've now spent a week living with a continuous glucose monitor (CGM). I've got a lot of thoughts, but it's safe to say that it's nothing short of amazing. CGM is a transformative technology for people with type 1 diabetes. I've been amazed by the accuracy of the sensor, how unobtrusive the design is, and how useful it has become to know my almost real-time glucose level. I wrote about it extensively in the first few days, which you can read here, here, and here. I'll try not to recap any of that in this post.

What I've learnt

It's been quite impressive to witness glucose levels spiking right after certain meals, and not others. While this has undauntedly been happening for years, it's the type of things you don't notice when you're taking intermittent glucose readings. It can be frustrating to watch the reading rise while knowing it might be an hour before it starts to fall again. I Tweeted saying that I'm considering not eating carbs again, mainly to avoid these spikes. The Tweet was hyperbole, but it has made me start to reconsider certain food choices. For example, a bowl of Weet-Bix in the morning will likely be replaced with toast as it doesn't have such a negative impact on my levels. It's also been positive to learn that the Dexcom CGM is accurate. It requires two manual calibrations per day, and most of these have given readings within 0.5 mmol/L of what the Dexcom is showing.

What's useful

Near real-time glucose data is truly transformative. It takes the guesswork out of managing diabetes and means I can choose to take action on my glucose levels at almost any time. Feeling a bit on the high side? A quick glance at the watch will either confirm or disprove this feeling, and means I can take action if necessary. It's also reassuring to be able to check my Apple Watch in the middle of the night and know I'll get an instant reading. There are times when the reading is not bad enough to have triggered an alarm, but the fact I'm half awake means I can glance at the reading and decide whether to give insulin. Despite my dislike for alarms on the phone during the day, they're great overnight. It's useful to me alerted overnight when my levels aren't great. While not a life-saving advantage, another is being able to leave the manual finger pricker at home. It's one less thing to carry and worry about, and one less thing to awkwardly take out and use in public. With a glance at my wrist, I can discreetly check my glucose level, and move on with life.

The sensor

The physical design of the sensor is bulky, but despite this, it is less noticeable in most situations than the smaller connection for the cannula of an insulin pump. It's easy to brush against, especially when placed in the stomach, but the tape holding it in place is strong, and it isn't easy to knock out. It's also more comfortable than the pump connection, as it's hardly noticeable when sleeping. I did mention last week that it is slightly uncomfortable in the car as it brushes against the seat belt, and also while running. The former is still true, the latter not so much. Perhaps it just took some getting used to. I've also spent some time swimming at the beach these last few days, and it hasn't been an issue there either.

Changing the sensor

Since the sensor had been in for a week, it was time to change it this morning. The Dexcom app alerted me to this by sending multiple notifications this morning informing me the sensor would cease to work at precisely 10:20 am, and that it was time to replace it. There are two parts to the CGM. There's the sensor itself which sits just underneath the skin. This needs to be removed every week and replaced. There's also the transmitter, which is connected to the sensor. It sits on the outside and is what crunches the data, and sends it back to the phone. The transmitter lasts about three months, so it is retained and connected to the next sensor. Pulling it out is straightforward enough. Once the tape around the edges is removed, the sensor slides out easily. Unlike an insulin pump cannula, it hasn't left much of a mark on the skin. Once it's out, the insertion process begins again (on the other side of the stomach). As was the case with the initial insertion, it requires two hours for the new sensor to "warm up." Once that time has elapsed, two consecutive manual blood glucose readings are needed to calibrate the sensor, and then the CGM is back in business.

Time away

It's not always practical to be within Bluetooth range of your phone. It's happened a few times during the last week that I've not been nearby my phone for a few hours at a time. The transmitter simply stores a history of these readings every five minutes as it otherwise would, and sends them across with the first sync after your phone and Dexcom establish a connection again. It works reliably, and to my knowledge hasn't missed a single reading all week.

That's been the first week. I am enjoying having access to the real-time data a CGM provides, and the way that it can help me make informed choices when trying to control my glucose levels. The Dexcom app now lives on the home screen of my phone, and as a complication on my Apple Watch face. It's incredibly useful to have access to this information, and Dexcom's app is impressively reliable.

If you're reading this and reside in Australia, I'd really appreciate it if you took a moment to fill out a form here. It sends a letter to your local member of parliament requesting funding for CGM technology for everyone in Australia who needs it, and not just those under 21 years of age 21 as is currently the case. I wrote a letter and sent it tonight, and would really appreciate support from others. If you've got any questions, I'd be more than happy to answer. You can ask me question via the form here, or on Twitter.

CGM Diaries: Day 3

Day three. I'm starting to get used to the CGM and learn more about how it's going to fit into my life. Today was a good test for it, as I was out of the house for an extended period.

This morning began with the first run I've been on since getting the Dexcom. Having an Apple Watch with a built-in 4G connection meant I didn't take my phone. A side effect of this was not being able to see my glucose levels while running, as the Dexcom connects to the phone but not to the Apple Watch. This will change as a result of watchOS 4 and changes to CoreBluetooth, but as I type this those features aren't supported yet.

The Dexcom sensor is comfortable enough for day-to-day activities. I haven't been able to feel it when at the gym, or when sleeping (which is what I was most concerned about. It has brushed up against the seatbelt in the car which can get uncomfortable, but it's no big deal. Where it's been most uncomfortable was on this morning's run. There's no other way to describe it other than it felt heavy. I've not felt this way about an insulin pump set, but the Dexcom is absolutely noticeable - and somewhat uncomfortable - while running. Again, not a huge issue, but it's something to note.

After years of being trained to complete regular "finger pricks" around lunch time, losing this mindset is taking some time. With the CGM, I can just eat. As strange as that sounds it's true. There's no longer a prerequisite to eating. Similarly, earlier today I found myself feeling a rise in my blood sugar. It was only minor, and usually I wouldn't worry about it. With the CGM, I was able to glance at my wrist, note that it was slightly higher than it should be, and give a dose of insulin through the pump. Not having to draw blood to get this information is helpful.

As I've mentioned previously, it is recommended to calibrate the Dexcom sensor with a manual "finger prick" every 12 hours. Being out of the house for so long today meant that these two calibrations would occur about 16 hours apart. That hasn't stopped the Dexcom app from sending me a notification every five minutes since 6 pm this evening reminding me to, "Enter BGL meter value." There seems to be no way to clear this without entering a false calibration - which I'm reluctant to do. On the positive side, these notifications aren't making audio alerts. Another positive is that all of the calibrations so far, except one, have been within +/-0.5mmol/L of the current Dexcom sensor reading. It's reassuring to know that it's accurate.

Today was the best day I've had so far for learning what makes the CGM useful. Highs and lows are unpreventable for someone with type 1 diabetes, but a CGM ensures you're on top of these swings and can help you recover from them faster. The highs won't be so high. Nor the lows so low. It's about the convenience of leaving your blood glucose monitor at home for the day. Did I mention? I haven't seen the thing since seven this morning - which is almost certainly the longest time I've spent away from one in over 16 years.

CGM Diaries: Day 2

I feel the need to begin by stating that this is in no way a complaint about the Dexcom G5 continuous blood glucose monitor and that I am extremely fortunate to be wearing such a device that gives me continuous, real-time data. It works well. Being able to see this data in real time will undoubtedly lead to better overall control. It's already saved a hypo. This post is simply what I've discovered when trying to write software based on the data provided by the Dexcom.

As a software developer, the thought of the data collected by the Dexcom is exciting. The Dexcom iOS app works well but isn't perfect. That's okay as long as there's access to raw data. In theory, access to all of the readings off of the sensor would allow me to build an entirely custom app for viewing real-time glucose readings on both iPhone and Apple Watch. This theoretical custom app would also include support for my preferred Apple Watch complication, as well as more granular control over notification and particularly the sounds they make. Another possibility would be a custom companion app, so that my parents could check in with my glucose level overnight. There are countless possibilities.


My first thought was to build something that pulls data from HealthKit and work with it that way. The Dexcom app currently writes data to HealthKit at a delay of an almost-perfect three hours. I addressed that in yesterday's post. While that is okay for an app that analyses longer-term data and trends, it isn't helpful for the projects I had in mind for today, with the goal being to use data that's as real-time as possible.

Dexcom API

A quick Google search this morning for "dexcom api" lead me to realise they do have a way for this data to be accessed by third parties, so I decided to play around with it. Many hours later (too many of which spent getting the authentication right), I wondered why the simple app I'd built to return data captured in a given period wasn't working. That's when I came across the following on the Dexcom website.


That explained why today's efforts were failing. It's disappointing as a developer, but I'm sure there are valid explanations as to why. As an FDA approved device, there are processes in place, and lots of regulation they have to adhere to before anything gets done. I can also confirm that this problem can't be addressed by simply changing your Dexcom account address to a United States one.

Where to from here

I'm a little deterred knowing that there isn't a reliable way to access real-time data from the Dexcom sensor. It doesn't matter much - the official Dexcom app works reliably and is enough for most people most of the time. A possible next step is to try and read data straight from the Dexcom sensor via Bluetooth. Some quick searching earlier this evening tells me this may be possible, but that's a project for another day.

CGM Diaries: Day 1

Following on from yesterday's post, today was CGM pickup day. This involved visiting the hospital to collect the first batch of supplies for the CGM, as well as set it up, and learn how to use it. This is more of a "first impressions" post, as it will take a while to be able to assess its usefulness in managing glucose levels overall.

The smaller box contains the Dexcom transmitter. The larger box contains the sets. 

The smaller box contains the Dexcom transmitter. The larger box contains the sets. 


Surprisingly simple, there are two parts to the Dexcom G5 CGM. There's the transmitter (shown above as the smaller of the two boxes), and the "sets" - the piece holding the cannula that attaches to your skin. The Dexcom iOS app has a straightforward onboarding process which ends with you linking the transmitter device by scanning a barcode on the back of the box.

This is used to fix the set to the body. It looks scarier than it is. 

This is used to fix the set to the body. It looks scarier than it is. 


Sensor insertion

Nearly lifetime of needles, nor over eight years of inserting cannulas for an insulin pump meant I felt calm when seeing the insertion device. It worked fine and fortunately didn't hurt, but that doesn't mean it wasn't intimidating. You can wear the set either in your stomach or on the back of your arm. I choose stomach, as I'm comfortable with putting the sets for my insulin pump there and figured it's a safe option. Once it's in, you clip the transmitter device into the plastic enclosure of the set where it sits for the next week. The transmitter lasts three months, while the sets are swapped every week.

Physical size

It's bulky. See the photos above. It's about twice as thick as the insulin pump sets I'm used to. It's not uncomfortable, but my biggest concern is that it'll be easy to knock out. For reference, sensor and the set combined are about a thick as the width of my index finger.


Getting started

Once the sensor is inserted, it takes two hours to warm up. From there, it asks you to do two manual blood glucose readings, one after the other, to calibrate the sensor. You input these into the Dexcom app, and from there it gives you your first reading. After that, it senses your glucose level every five minutes and sends that reading back to the phone. Interestingly enough, Dexcom makes the only CGM sensor that is accurate enough to trust as a replacement for blood readings. Despite that, two manual blood readings are still required per day to keep the sensor calibrated.


The iOS app

The iOS app is fairly simplistic. It shows you your current reading, along with an arrow indicating the trend of the readings. It's worth pointing out that, for those who care, the iOS app is still running at iPhone 5 resolution - not yet updated for 4.7" and 5.5" displays, let alone the 5.8" screen size of iPhone X.


Notification alerts

The Dexcom app offers real-time notification alerts for glucose levels that are either too high or too low. These somehow bypass silent mode, do not disturb mode, and the system volume level on iOS. I've received a few of these notifications already (don't judge - the CGM should help reduce the number of poor readings I have) and it's safe to say they're extremely annoying. My phone has stayed in silent mode from the second I got a smartwatch four years ago, and I don't like the fact that notifications are making noise again. More granular control over this would be appreciated. It may be useful overnight to alert me to hypo (low) glucose levels, but I wish there was a way to silence them. A tap on the wrist from the Apple Watch is all I need to see an alert and action it.


Apple Watch complications

For a variety of reasons that I won't go into here, I've used the "Simple" Apple Watch face (shown in the first image above) almost every day for two years. I am also of the opinion that having an Apple Watch and a Dexcom CGM means you are obliged to use the complication to get your BGL at a glance. Unfortunately, the Dexcom complication doesn't support the Simple watch face. The closest alternative is "Utility" - shown in the second image above. It means I lose having my step count on the watch face, as there's space for one less complication, but I think that's a worthwhile tradeoff for now. The only watch face that allows me to have the Dexcom complication as well the other complications I'm used to is "Modular" (third image above) but I'm not a huge fan of that one. I might use it as my overnight sleeping watch face but will try to avoid it during the day. It's too early to tell for sure, but the complication seems to update every 10 minutes, which lines up with every second reading recorded by the Dexcom sensor, so it's frequent enough. Occasionally I have noticed that it fails to update, and instead shows three dashes "---" (shown in the fourth screenshot above).


HealthKit sync

Another reason for choosing the Dexcom CGM, other than it being the most accurate, was that it's the only one to sync to a phone. If you have an iPhone, not only does it sync to the Dexcom app on your phone, but also optionally can write the data to HealthKit. This is perhaps the most strange thing I've discovered about the Dexcom so far. The Health screen reads, "Dexcom G5 Mobile information posts to Health with a three hour delay." My first thought upon reading this was that it only writes to HealthKit every three hours - i.e. eight times per day, maybe in an attempt to save battery?


As far as I've been able to discover, the Dexcom app writes to HealthKit every five minutes but writes the reading from three hours ago. I'm not sure why this is, and I'll continue to try and find out why. It feels arbitrary. As can be seen in the above screenshots, a reading recorded at 2:29 pm was written to Health at 5:29 pm, and a reading from 2:34 pm was written to health at 5:34 pm.

Wrapping up

I've still got a lot to learn about the Dexcom CGM. The best way to use it, how to interpret results and trends, and how it fits into my life all remains to be seen. Some sections may have sounded a bit like a complaint, but I assure you it's just an expression of my initial impressions. I am in no way complaining about the bulkiness of it, the missing "Simple" complication, nor HealthKit sync. I'm in a very fortunate position to have a CGM, and these are simply the things that have stuck out to me that I wanted to share after six hours of use.

CGM Diaries: Day 0

Continuous glucose monitoring, so I'm told, is life-changing technology for people with type 1 diabetes. I spoke to someone last week who told me her HbA1c result (essentially an average blood glucose level over three months) was down to 6.1 mmol/L from 8.1 mmol/L. That's 110 mg/dL and 146 mg/dL respectively. I've heard similar results from others. That is no small improvement, and on top of feeling better in day-to-day life, it dramatically reduces the chance of developing complications from diabetes later in life.

The range of continuous glucose monitors (CGM's) available now are all invasive, meaning they puncture the skin to work. I've written before about what non-invasive blood glucose monitoring would mean for people both with and without diabetes. The idea excites me beyond belief. However, it may still be years away, and for now, the choice is whether someone sees the benefits of CGM enough to accept the tradeoffs. If someone with diabetes is also wearing an insulin pump, expecting them to live with another cannula in their body (generally either in the stomach or back of the arm) may be a dealbreaker.

Inconvenience aside, the cost is probably the most significant factor preventing more people from wearing the current CGM's available. Diabetes Australia says the cost of CGM, including sensors, is around $5000 per year. That's no small amount of money and puts the possibility of wearing a CGM out of reach for many people.

With some notable exceptions, it's nice when an elected Government follows through with an election promise. The current Australian Federal Government committed to wholly subsidising the cost of CGM devices for Australian's with type 1 diabetes who are under the age of 21. This is thanks to tireless lobbying from the Danii Foundation. Fortunately, I am eligible for the subsidy for the next 11 months.

After speaking to the doctor at my last checkup, she agreed a CGM would be beneficial, and I'm scheduled to have have one set up tomorrow (4th December, 2017). The CGM model is the Dexcom G5, which was recommended as it's the most accurate, and can replace all but two "finger-pricks" per day for checking blood sugar. It also transmits data via Bluetooth, and as a result, will sync the data to HealthKit on my iPhone and Apple Watch. The ability to store this data reliably for future reference is a definite advantage.

This is an exciting time when the possibilities are considered. Improved control and management of blood glucose levels will almost certainly lead to advantages I can't even think of at the moment. Because of the integration with iPhone, Apple Watch, and HealthKit, I also plan on writing about it on this blog. Those posts will likely focus on the integration between the Dexcom sensor and this technology, as opposed to how CGM is affecting the management of diabetes and my blood glucose levels. On the other hand, a bulky sensor may be deemed not worth the tradeoff. I'm curious to find out.

While I've got your attention on this topic, we're still raising money for JDRF. Their research leads the way in working towards a cure for type 1 diabetes. Any amount you've got to give goes is greatly appreciated. You can read more, and donate here.

Experimenting with App Store Search Ads

What are Search Ads

App Store Search Ads have been around since October of 2016. Apple describes them as, "An efficient and easy way to promote your app at the top of relevant App Store search results." They were only shown in the U.S. App Store up until April of 2017 when they were introduced to the UK, Australia, and New Zealand. Since then, they've been added to the Canadian, Mexican, and Swiss App Store, bringing the total number of App Store markets with Search Ads to 7.

Here's how Petty's Search Ads look on the App Store

Here's how Petty's Search Ads look on the App Store



I make Petty which is an app for iOS and Android that helps New South Wales (NSW) drivers find affordable petrol nearby. Petty came about because I thought the NSW Government's petrol price dataset looked cool and knew that it was the most accurate one around. At that stage the best tool one could use to view prices from the same database was to use a website, called FuelCheck, run by the NSW State Government. I thought the idea of a decent native app, as opposed to having to visit frequently was appealing.


About a month ago the NSW Government launched FuelCheck in the form of a mobile app for iOS and Android. Its release was inevitable. As best I could tell, it "soft launched" on the App Store and Google Play a few weeks prior, but the hard launch date was the 11th of October, 2017.

The first I heard about the launch was in a Tweet by the NSW Premier, Gladys Berejiklian.

Naturally, there was a lot of attention given to FuelCheck. Petty and FuelCheck are two different apps and Petty has things to offer that FuelCheck doesn't, including showing the time prices were last updated at nearby stations, an Apple Watch app, notification centre "Today" widget, and - as I Tweeted at the time - assurance that you aren't using something made by the government.

It was half-jokingly suggested by a friend of the blog (who didn't get a say in their friend of the blog status) Pat Murray to buy Search Ads on FuelCheck's search terms. That evening, I set up an ad campaign and did precisely that.

Search Ads for Petty

To provide some context, Petty is a free app with optional in-app purchases (IAP). Most of its (insignificant) revenue comes from the IAP to unlock "premium" mode - removing ads and getting extra features. The rest is from an optional tip jar. The money made on display ads is so small it might as well be a rounding error. It's worth noting that any money or dollar figures mentioned are in Australian dollars (which, for the fun-fact enthusiasts among you, is the fifth most traded currency. The ads for Petty were only shown to customers who have not previously downloaded the app, and those who were located in Sydney, Australia.

The ad campaign

I hadn't run a Search Ads campaign since Apple gave $100 in free credit at launch to every developer encouraging them to try it out. Early on it was established that I was only willing to throw $30 at the ad campaign and that the ads were going to run on the iOS App Store only despite there also being an Android version of Petty. $30 sounded like a nice number - small enough that it didn't matter if it resulted in zero additional revenue, but enough to gather some data with. I expected the ad campaign to be over sooner than it was, as maximum daily spend was set to $7.00. It took almost a full month for the $30 budget to be exhausted.

Here's a breakdown of the spend per week

Here's a breakdown of the spend per week


The ad campaign finished with 927 ad impressions, and 168 taps on those ads which converted into 101 installations of the app. This gives a tap through rate (TTR) of 18.12%, and a conversion rate (CR) of 60.12%. I'm particularly impressed with the CR which shows that a majority of people pressing the ad were interested, and went on to download it. There isn't much public data I can compare this to, but it's slightly higher than what early results for Search Ads showed - a good sign overall.


The most important metric when running an online ad campaign for an app is the cost per acquisition (CPA) of a customer. This campaign saw an average CPA of $0.30, which is lower than I expected, even for a free app.

CPA is an important metric because it tells you whether or not your campaign is worthwhile. The marginal cost of an additional customer of your app is almost zero. In a simplistic model, take average cost per customer (such as data, and server expenses) away from average revenue per customer, and you're left with a bit more than the amount you can afford to spend to acquire a new customer.

These keywords were the only three to see meaningful results 

These keywords were the only three to see meaningful results 


The ad campaign began with only a few keywords: petrol prices, fuelcheck, fuel check, and fuel watch. Throughout the month the ads were running, I increased the number of keywords that ads were running against. At the end of the campaign, only three keywords generated more than 100 impressions, and more than 5 installs each. They were: petrol prices, petrol, and fuel watch.


Interestingly enough, ads for Petty did not score so much as a single impression on either "fuelcheck" or "fuel check" - the two most important keywords set. I know the Government themselves were bidding directly on that keyword, as were 7-Eleven with their fuel app, and their default CPT bid was likely a lot higher than the $2.00 I set for Petty ads.

After a few days, "fuel watch" was trending, but not, "fuelcheck."

After a few days, "fuel watch" was trending, but not, "fuelcheck."


That means these ads weren't shown to people who were visiting the App Store and searching for FuelCheck directly. The ads likely were shown to people who had heard about the Government's app but weren't sure what it was called. It is interesting that a trending search term on the App Store only three days after the launch of FuelCheck was "fuel watch." It trended for about 24 hours, and as you can see from the data above, that was the single most popular search term for Petty's ads during the campaign.

Wrapping up

There are a few key points to take away from this experiment. Search Ads on the App Store work. Unlike Facebook or Google Search ads, where you pay $20 only to get the email, "Your campaign has ended, and your page has ONE new like!" Your mileage may vary, but the $30 I spent resulted in just over 100 new customers, at a cost to me of $0.30 per customer. As far as CPA goes, it was an effective means of advertising. Another thing to take out of this is that you don't need to spend a lot of money to see results. $30 was enough in this instance, and it lasted nearly a month. $20 probably would've been just as fine. Play around with different ads and keywords to figure out what works for you and your app. Run an inexpensive campaign to figure out what your CPA is, and go from there. I also learnt it's difficult to outbid the "big guys." They're almost certainly going to bid higher on the popular terms, such as "FuelCheck," so it's important to find some moderately popular, niche keywords ("fuel watch," in this case) to see results.

There's more data that you can take from a Search Ads campaign and more detailed analysis to be done on the data than I've written here. I hope this provides a good overview of what to expect when you run a Search Ads campaign and how it can be an inexpensive way to attract more customers to download your app.

iPad Productivity at Uni

I think a lot about productivity, and how to be more efficient when working. The most prominent challenge for me relates to University work. I've always found it easier to achieve "deeper work" (more extended, uninterrupted periods of working time) when working on side projects than when doing things for Uni. This will be a short post, but I'm writing it here as opposed to on Twitter because Twitter dot com doesn't need another Tweetstorm.

I'm in an arguably lousy habit of working on side projects while at Uni. It begins when I pull my laptop out on the bus, start working on something, and then continue that when I get to Uni. Sometimes this means spending the first few minutes of a class wrapping up said work. Later, during the next break between classes, I'll continue. Procrastiworking (doing productive work that just isn't the work you should be doing at present) isn't inherently bad; however, it does mean I'm not working on Uni work at Uni outside of time in class. I tend to do most of my Uni work at home, and I get it all done there without issue, so I guess I'm trying to complete more while at Uni.

In a hypothetical world where I only take an iPad to Uni (as opposed to the MacBook Pro I carry now), I would probably be able to complete about 80% of the work I need to. The remainder (subjects involving programming assignments) would have to be done at home, but seeing as though I already do most of my Uni work at home that would be fine. I could take notes with an Apple Pencil, access all of my folders and documents through the OneDrive (or Files) app, and browse the Internet to conduct research. iOS on the iPad has come a long way recently and is certainly up for the challenge.

In this hypothetical world, I can see things going one of two ways. Either the iPad helps with focus at Uni, or it doesn't. Sounds obvious, right? There's no decent IDE for me to open and start working on a project with, so in theory, I'd have to do Uni work. This would also apply to time on the bus. If I pulled out my computer - in this case, an iPad - the only productive tasks I could complete would be work for Uni. The second possible outcome is that I just find another way to be distracted. More Twitter perhaps? Goodness knows I need less of that in my life. If the latter proved to be correct, it would be a net loss of productivity. There would be no procrastiworking, but no additional "real work" completed either.

No conclusion has been reached here. It was just a fun little thought experiment on my last day of classes for the year. I almost certainly won't buy an iPad for Uni, but it's fun to think about different ways of working.

Even the act of writing this was a form of procrastiworking between classes. Bring on exams. 🙃

Petty 1.1

Today, the first major update to Petty (v1.1) goes live today on both the App Store and Google Play Store. It introduces a search feature - you can now search for any station in NSW, view its price list, and get directions. There's also a new Today Widget, and Apple Watch app for iOS users. The update also intentionally coincides with the public release of iOS 11, as it adopts some of the new iOS 11 design trends. Mainly the large title/navigation bars.

Side by side.png

I was reluctant to adopt the new navigation bars, as I'm not a fan of the look, but after seeing iPhone X and how they blend in on that display, I became more convinced it's the right choice. It's also just a good general rule to follow the design patterns of the platform you're designing for.


Search, it's here! Ideally, this would've made it into v1.0, but I ran out of time. It can't be more straightforward to use. There's a search bar up the top of the app on iOS, or a search button on Android. Use it to search for petrol stations in NSW. Analyse the results to your heart's content.

Today Widget.png

If you're a widget person, there's now one for you on iOS. Petrol prices, they're everywhere. Why not, right? You can access the expanded version from Notification Centre, which shows you the closest station and its full price list. The smaller version, which can be accessed by 3D Touch-ing the Petty app icon on your home screen, just shows you the closest station, its distance, and address. There's no room for a price list there.

Apple Watch.png

Do you wear a shiny computer on your wrist? If so, this update is for you. Petty now has an Apple Watch app. It will show you the nearest station, its distance, and its address. You also get the option to view the station on a map. That is all. I can't label the station, give it any other metadata, nor can you interact with the map. If you want directions, you'll be booted to the Maps app on the Watch. Unfortunately, that is a current limitation of showing maps on the Apple Watch. Hopefully, it changes in the future. This feature may be especially useful if you've got, or are getting, a fancy new Apple Watch with a red circle on the crown. If you do, that means you've got a watch with a cellular connection. You don't even need your phone to find petrol - how cool is that? It's almost as though I preempted the lack of need to carry your phone around while driving and running errands. I am, however, definitely overestimating how keen you are to know petrol prices. Moving on.

Actually, there's nothing to move on to. That's it for this update. I hope you like it. There will be more to come in the future. Let me know what you think.

Like the look of Petty? Download it here for free.

If you're a fan, maybe tell your friends, and then drop a 5-star review for Petty on the App Store and Google Play, it's appreciated. For more updates, you can follow me on Twitter - @zachsimone - where I'll sometimes post previews and sneak peeks.

Social media 2.png

HealthFace 2.0 Overview

This week's update to HealthFace is the first since May. On the same day that the last update hit the App Store, I had a Twitter exchange with the developer, Quentin. As a result of this exchange, I became a regular user of the app.

The new views in HealthFace 2.0 for data entry and extension customisation

The new views in HealthFace 2.0 for data entry and extension customisation

HealthFace is a health app for iPhone and Apple Watch. Version 1.0 was all about displaying data from Apple's Health app on the watch face of an Apple Watch via the complications feature. Version 2.0 came out earlier this week and turned the app into so much more. It's now my go-to app for inputting data into the Health database on my phone. Manually inputting data through the Health interface is clunky and unpleasant. HealthFace solves this.

The new Apple Watch input feature 

The new Apple Watch input feature 


My use of HealthFace begun by setting up a complication on the face of my Apple Watch to display a 7-day average of my Blood Glucose level. To this day, I have the same complication in the same spot on the same watch face on my Apple Watch. Around the same time, I began saving every BGL reading taken into the Health app on my iPhone. I have type 1 diabetes, and this information is useful at a glance to learn how well I've controlled my blood sugar for the last week. If HealthFace just offered custom complications, that would make it a good enough app to keep around. Over the last few months, I've excitedly watched it transform into something that takes it from a niche app and turns it into something for everyone - with or without an Apple Watch.

Using the Today Widget is a quick way to input data

Using the Today Widget is a quick way to input data


Everyone has a health metric they need to track. For some, it's blood pressure, or volume of water consumed. For others, it's simply daily weight measurements. By keeping on top of these metrics, it keeps them front of mind, and can subconsciously encourage better habits. Every iPhone ships with the Health app which can serve as a fantastic collection of all kinds of health data. If there's a health-related metric you want to keep on top of, there's an excellent chance you can record that data in the Health app. HealthFace 2.0 enables easy entry of this data into your health database. You can enter it from the main iOS app, from a widget on iOS (meaning you can enter data when your phone is locked), and on the Apple Watch. I use all three methods depending on the situation I'm in, but I find input via Apple Watch particularly useful.

The input interface is unique yet intuitive 

The input interface is unique yet intuitive 


Above is an example of the custom input mechanism built into the iOS app. There's a stepper to increment the value, and also a touchpad which can be used to "slide" the values. It's faster and more intuitive than using the Health app to save data.

The app is fully customisable. You can choose types of data and set them as "favourites." These favourites then appear across the iOS app, widget, and watch app, or you can choose different types of data for each app extension. e.g. Blood glucose and blood pressure on the widget, but water intake and body temperature on the Apple Watch. Everything Quentin's added to this update aids the goal of quickly adding data to health.

Interestingly enough, the "Complications" tab - which used to be shown on the first launch - is in the fourth spot as of v2.0. This illustrates the focus of the app has changed with this update, emphasising the other features of the app.

New icons!

New icons!


As a nice bonus, there are some new custom app icons. HealthFace is no longer an app for just people with an Apple Watch. It's an app for everyone with an iPhone. It's available on the App Store for A$2.99 (US$1.99).