Apple’s Unusual Situation with its Struggling HomePod

HomePod Siri light

Apple, as it always does, created a unique vision for the HomePod. The device is cool, but the product concept may have been off the mark. Now, Apple will adjust.

Apple HomePod and Amazon Echo
They look the same, but are very different in concept.

The Particle Debris article of the week is from Mark Gurman at Bloomberg. He explains how It appears that Apple may not have sold as many HomePods as hoped.

Of course, no company ever sells a product in the quantity it desires, but the real questions are: how were the predictions generated and why hasn’t the product met expectations?

I can sympathize with those at Apple tasked with sales projections. People who do that want to be seen as insightful, enthusiastic proponents of the corporation’s product. Low estimates are met with scorn while higher estimates are met with nodding approval. But no matter the predictions, its up to the sales force to meet them—or take heat.


By all accounts, Apple was caught off guard by the Amazon Echo. The scope of the effort, in time and resources, was underestimated. It may have been that, as the concept for the HomePod evolved, its design was dictated by privacy policies and the state-of-the-art with Siri. In any case, when the first Echo shipped, it clearly had a different kind of vision than Apple’s. The Echo’s design makes perfect sense; Amazon needs to understand and enable its customers who frequently shop for its wealth of products. Apple was thinking different.

I think Apple surmised that it was ill-positioned to compete with its HomePod, already in progress, and elected not to. It would be easy to surmise that customers would love a great music product, and that created a convenient rationalization that head-to-head competition would be unnecessary. Time will tell if that was a bad assumption.

Electing Not to Compete

Most tech giants, seeing the enormous success of the Amazon Echo family and the Google Home family would mildly panic. Plans would be put formulated to aggressively jump into the fray. Apple, in fact, doesn’t mind doing that. The company sized up the MP3 player market and produced the iPod. It sized up the smartwatch market and produced the Apple Watch. Both these products, thanks to their superior design and integration with Apple’s existing family of products, went on to dominate. Perhaps that’s the plan for the HomePod.

[Here’s How I Finally Accepted HomeKit is a Raging S*#t Storm ]

Apple has a history of tweaking a product’s initial, sound design until its virtues become more broadly appreciated. There are rumors that Apple will introduce a lower cost HomePod as part of that strategy. Apple competes in its own way and its own time.

HomePod’s Future

In the final analysis, however, Apple produced what seemed to be an expensive, me-too product with serious shortfalls. Many consumers were mystified by its design and annoyed to discover unexpected limitations. The HomePod declined to compete directly with popular smart speakers because it could not. And did not plan to do so.

Now, Apple is about the business of deciding how to make the HomePod’s virtues more widely recognized, seriously change its design, lower the cost, expand the family, or rethink the product completely. Apple has time to produce a better smart speaker, a low-cost sibling to the HomePod, that could outclass the competition. If the company elects not to do that, the plight of the HomePod itself, however resuscitated and reimagined, will be interesting to watch.

Next Page: The News Debris for the week of April 9th. Are autonomous cars doomed by pollyanna expectations?

11 thoughts on “Apple’s Unusual Situation with its Struggling HomePod

  • Ok, something many tech writers have failed to notice in the past: Many of Apple’s most popular products ever had slow sales initially -like the iPhone🙌🏻. This is probably due to Apple’s strong and popular ecosystem, which boils down to the fact that basically, mom, dad, brother and sister all have iPhone’s, iPad’s, AppleWatch’s, maybe a family iMac, MacBooks etc etc. Calculating that would easily account to way above $20 000 per family every 4 years or so… With that said, many consumers as wealthy/snobbish as they may seem, don’t necessarily have cash on hand just to walk into an Apple Store and buy the next big Apple product, especially just after they already upgraded their iPhone’s to the latest iPhoneX that came out JUST before the HomePod.
    So that means having spent waay more than any Android consumer who is more likely to buy more cheaper gadgets, therefore the Android user who can easily buy a handset for for even less than a $100 dollars (and still falls under the same middle class income as the Apple user) therefore still on average has a couple of hundred dollars more left to spend on other gadgets such as Amazon Eco’s and Google home speakers.. Android users go for quantity, while Apple users go for quality. In other words, the Apple user prefers the cake later…

  • Recently playing an online game and a commentor was complaining about “everyone camping out” as the team was losing. When we came out of hiding, turned the tide and won the match, I explained: “Sometimes camping out can be a strategy.” – holding forces in reserve. I don’t think this has been Apple’s strategy. Half baked projects and knee jerk reactions. Focus on better sound than the competition but ignore the “smart” home with a consistently dumb assistant. And push the limit on pricing, see what the market will bear. Thinner, lighter, a new color, a notch, more speakers; for the same price or more. I’d be shocked if they are holding anything in reserve. The recent leaking memo may be more about embarrasment of nothing to reveal than any actual revelations. Who was Phil Schiller refering to when he said, “Can’t innovate anymore….”? Must have been some scuttlebutt that Apple’s best days might be behind them.

  • John, ironic you would have an early iMac commercial, because vehicles that age are still on the road. 🙂

    The tech industry will need to come to grips with a product that has a much longer service life than, say, a smart phone. When did the Mac G3 last have an OS security update? What happens when a 10+ year old self driving car needs a safety-critical software update? “Time to upgrade, you got your use out of it” is the wrong answer, and isn’t going to fly with government regulators.

    Hopefully in a decade the serious bugs will be ironed out, but there is going to be a learning process.

  • @[email protected]:

    I think we’re in agreement.

    As I argued above, an ideal system is one in which the individual chooses the route, the car AI executes that choice but the central AI manages traffic flow, not unlike air traffic control manages air traffic. The individual should be free to opt for any change they wish, such as changing the route or stopping for a coffee – just like we do today with no restrictions. Central AI controls the traffic. Auto AI the automobile and the driver decides when and where to go.

  • I think a few more software updates can make the HomePod pretty awesome. It definitely has the best sound of any of the speakers out there. More Siri support, Airplay 2 needs to be done already and also third party support for Pandora and they will easily have a hit on there hands.

  • John:

    I think one of the important services, albeit with a delayed tangible benefit to the readership, is your continued focus on the self-driving cars, in addition to robotics and AI. But, since you brought up self driving cars, let’s talk about this.

    To begin with, let me concede that the Wired article does indeed ask important questions. Where I diverge is that I think these are not merely the wrong questions, but that they reflect a conceptual shortfall and failure of imagination that is not confined to Wired, but to a great many commenters and analysts. I’ll be the first to acknowledge the moxie if not the unmitigated gall of a non-professional in this field to criticise the thinking of those for whom this is their bread and butter; but shamelessly let me soldier on.

    I’ve cited Kenneth Dolbeare before, but in one of his early works that I had to read in university (Political Change in the United States, if memory serves) he pointed out that most revolutions fail, and that a principal reason is a failure of imagination in the scope of that revolution. Most end up reinventing and resuming the very systems that they sought to overthrow. It is exceedingly difficult for most people to truly think outside of the parameters of what they have seen or experienced, and thus they recreate the same systems that they sought to overthrow with modest, if not cosmetic changes. At best, this is evolution, but not revolution. Ultimately, they are saddled with many if not most of the same problems as before.

    In that vein, the self-driving car is the wrong solution to a pending crisis; efficient, clean and above all safe transport for society. The concept of a auto-piloted car, driven by AI as a self-piloted and self-contained system, is a mere variation on an archaic them of individualised transport. Just as the donkey and horse were replaced with a machine, now we propose to swap the human for machine learning algorithms. Despite it’s potential for improved safety, it does little for cleanness and virtually nothing for efficiency. This too is simply an evolution, not a revolution. It moves the needle but little, and will leave us with many, if not most, of the same problems.

    What would constitute a revolution? A wholesale reimagining of transport that, were we to start de novo today, would leverage our current technological skillset, and take it to the next level – a true moonshot. I’m confident that we would take a systems analysis approach to the problem, which is transport writ large; addressing the core modern exigencies of safety, security, environmental impact per unit of transport (sustainability) and efficiency (economic impact and therefore sustainability).

    Where would we start? We would not start with the vehicle. We’d start with the grid. In systems design and development, organisation and growth is organic; meaning you start with a basic body plan, not unlike a human body, identify its needs and then elaborate the organ systems needed to fulfil those needs, within the context of that body plan. And the first organ you would develop is the one that nature does, the brain, which will help drive and mould the rest of the design. The brain for a transport system is AI, not a vehicle simply equipped with its own isolated AI. The latter would be tantamount to designing a human body and giving each organ system or blood cell its own brain. You would not want these systems to think independently, rather to be coordinated as an integral system by one central organ. The same applies to transportation or any other system.

    Why would you not simply invest each little organ or organette, like a car, with its own intelligence as the answer? Because no system is ever stronger than its weakest component, and in the case of transportation, that would be the human. And unless vehicles are equipped with compulsory AI guidance with no human over-ride, a fundamentally unsafe, insecure and unpalatable idea, the human can and will over-ride the system at the worst of times (see Murphy’s Law). Mind you, this does not mean that there is no role for AI in the vehicle. Indeed, there is. More on that in below.

    How would development proceed? AI would control, meaning regulate and harmonise, the entire transportation system; just as the brain regulates and harmonises all of the organs of the human body. This means that it must be connected to and integrated with all core systems of the transportation grid, including roads (constructed with ‘smart’ components), traffic lights, bridges, tunnels, toll booths, exits, security cameras, etc. All need to be under AI regulation, including any vehicle actively engaged in that system. This means that when your car is in your private garage, it’s under your control. However, the moment you take it onto a public surface, it yields to central regulation, not unlike a pilot of an airliner submits to traffic controllers, thus harmonising all flights and keeping all passengers safe.

    The owner decides where they are going (including the route), but AI harmonises the flow of traffic in which the owner moves in getting there. Central AI has no control or say in any destination or route choice, or changes along the way of the human, but the human has no potentially detrimental input on vehicular performance on the public road. If the human decides to change the route, return home, take a break or whatever, they input this to the auto’s AI, who then coordinates with the central AI to make it so. Safely and securely. The human should have control only in the event of emergency or the unexpected, or on smaller service or private roads and private property, unless they prefer to handover to the auto’s AI.

    The implications for law enforcement, surveillance, and potential sabotage are as obvious as are the potential benefits to security, both personal and public, efficiency, environmental impact and the like if done with the right protocols and hardened support systems.

    This is what revolution would look like; as daunting and scary as it is potentially rewarding and beneficial as has been any revolution or brave new world into which we have ever ventured.

    And therein lies the problem.

    1. Central AI has no control or say in any destination or route choice, or changes along the way of the human, but the human has no potentially detrimental input on vehicular performance on the public road. If the human decides to change the route, return home, take a break or whatever, they input this to the auto’s AI, who then coordinates with the central AI to make it so. Safely and securely. The human should have control only in the event of emergency or the unexpected, or on smaller service or private roads and private property, unless they prefer to handover to the auto’s AI.

      Indeed, therein lies the problem. The assumption that the Central AI “has no control or say” in the destination or route choice is a very poor one. The temptation to meddle, control, dictate, or otherwise ‘tell others what to do’ in the envisaged automated road system is way too strong. We have enough problems with overreach in surveillance and demand for encrypted data from law enforcement (FBI vs Apple for example). A centralized AI for road management would be far too tempting a target.

      Yes, a centralized system may be the most efficient, but sometimes efficiency has to take a back seat to other, more important issues.

  • As a happy owner of 3 HomePods, I am a bit at odds with your article. I have two HomePods in my living room that are the sound output for an AppleTV. Yes, I am running the beta software for this to work. When I first got my first HP it needed to be almost at max volume for my 77 year old ears to enjoy the sound in a room that is 18’ x 24’ with a 15’ vaulted ceiling. The third unit is in the bedroom (recent birthday present for my wife) where she can listen to music as she dresses in the morning and we went to sleep with thunderstorms playing last night. She really enjoys the sound of rain. So, software is improving the experience of HomePods at an acceptable pace.

  • Very good points about autonomous cars. I keep thinking of the early days of aviation, when, yes even commercial planes often had trouble and killed people. But the public was able to see the future and put up with the risk. Remember Glenn Miller, Will Rogers, and Carole Lombard were all killed in plane crashes but people didn’t panic and stop flying. However I wonder if people’s expectations are too high for autonomous vehicles, pollyannaish as you said. They expect perfection even as the systems are developed.

    Today perfection is the only standard. Police must always get the bad guy and no mistakes are tolerated. Surgeons must always be Hawkeye Pierce, and do everything perfectly with a twinkle in their eye. Athletes must, for what they earn, always skink the shot and hit the home run. No one in public view can ever make a mistake, or fail in some human way. Technology must be perfect or it will be abandoned. Better is not acceptable. Batteries must not wear out. Phones must not bend even if I sit my fat @$$ on it. Car accelerators must never misbehave even if I stuff an aftermarket floor mat under it. Even if autonomous cars kill half as many people each year than manual drivers, that’s half too many. They have to be perfect now and forever.. Jakob Brownowski called it a “a tremendous loss of nerve” in the west leading to a retreat from moving forward. I wonder if he was right.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.