No, I’m not talking about the strategy in which Apple rolls out a smart speaker years behind its competitors — the one that has left former Apple employees and the world at large confused as to why the wealthiest, most design-centric tech company in the world took so long to unveil HomePod.

I’m talking about the other Siri strategy — the one to gradually roll out new ways for iOS developers to connect their apps with Siri.

Siri is nearly six years old, the first popular intelligent assistant of its kind in a smartphone. But Apple is a company that gained its status not by being first to market, but by producing a superior product.

As with the Apple Watch and iPhone, years from now Apple can tell us all how brilliant it was to quietly apply the lessons it learned from the early days of the App Store — when the total number of apps did not ensure the delivery of value. Instead of going for a large number of apps, Siri would do well to go deep and help users solve their biggest problems.

Getting into integrations

Alexa, Siri, and the like are plenty smart on their own, but it’s the third-party integrations that continually wow users and remind them why they should get a phone with Siri inside or meeting hardware that talks to Alexa.

Using Sirikit today, developers can accomplish a limited range of tasks. Outside the core list of things Siri can do with default iOS apps, like turn on the lights or make a phone call, the assistant can connect with WhatsApp to send messages or with the MyWorkout app to start a workout.

Ask Siri “What can you do?” and a list of apps that connect with the intelligent assistant emerges. In my case, only 16 apps, including Uber, Square Cash, and Twitter, showed up.

It’s unclear just how many apps are connected to Siri today. VentureBeat reached out to Apple for an answer to this question but received no response.

Whatever the exact number is, it appears to be small potatoes next to the more than 15,000 skills currently available on Alexa. But that number hides a poorly kept secret: All app stores, including Apple’s App Store, begin with a long tail of crap.

There are definitely some great voice apps available today, and Amazon is investing tens of millions in the creation of an Alexa ecosystem, but the Alexa Skills Store is starting from scratch. It’s the fart app scenario all over again, and there are certainly parts of the Alexa Skills Store that stink. A WurstWetter skill accomplishes the pointless task of telling you about bad weather in the world, while the one-star Fart skill is, well, exactly what you’d imagine.

Personalization makes bots useful

I’ve never wanted an intelligent assistant that does everything. My biggest desire has always been to personalize, to incorporate the list of 20 to 30 tasks that I do all the time and that are of vital importance to my life. That’s probably why people have used IFTTT to make more than 750,000 personalizations for Google Assistant and Alexa.

The list of tasks that an assistant can be most helpful with is going to be different for everyone, but instead of intelligent assistants building for Internet of Things (IoT) and customer service, areas where tech giants have tended to focus, they should spend more time asking users what they want.

Even at this stage, Siri, Samsung’s Bixby, and Google Assistant are fast becoming more than assistants you chat with to complete tasks. Scheduled for release this fall, iOS 11 will bring personalization powered by Siri to Safari, Maps, and other native apps. Google Assistant plans to do the same. As these assistants on smartphones learn about me, they should help me identify the list of things I can do with voice to lighten my daily load. It’s this sort of personalization that will define the fourth industrial revolution and could give Apple another chance to prove Siri’s value to the world.

iOS 11 will incorporate banking and personal finance apps and will deliver information about things like account balances and the amount spent in the past week or month. Siri is also learning how to chat with productivity apps, like OmniFocus. Instead of just being able to add items to a reminder list, Siri will soon be able to edit previously existing to-do-list items and create lists. QR code scans and payments are also on the way.

An app that helps people pay their bills and stay on top of personal finances is no fart joke. Some popular and interesting fintech bots are trying to fill this demand, including Digit, Trim, and Kasisto’s KAI. Personal finance is arguably one of the only examples of successful bot use cases discovered thus far. Apple would be smart to take a page from (or acquire) these sorts of companies, because today Kasisto can answer about 1,000 or more questions about personal finances and financial literacy.

Like the personalization of digital health that AI and wearables are bringing about, Siri could make specialized financial plans for teenagers going to college or adults without a 401(k), or one simply based on your financial goals and previous spending habits.

Speak up and win, Siri

I want my assistant to know me better than I know myself and to then take a mountain of data — gleaned both from me and from the aggregate group of 375 million iOS devices with Siri inside — and tell me how I can do things better. It should not only know what I need to do but learn what sort of behavioral nudges effectively keep me on track.

I want my assistant to help me pay my bills, manage projects, address prioritized tasks when I get to work (Siri in iOS 11 will factor in your location if you say “Remind me to print the presentation when I get to work”), and make sure I’m home for dinner on time.

If I’m late and I deviate from the fastest route, speak up. If I haven’t moved for an hour, send me a message. Given all the clinical health work Apple has been doing lately, Siri should know my medical history and incorporate computer vision to tell me which items on a menu meet my dietary requirements.

Going deep may be increasingly important for Apple as competition from the biggest tech companies in the world continues to heat up. Indications from Samsung, Baidu, Alibaba, and Line in recent weeks suggest they are looking into building their own smart speakers.

Siri is the original intelligent assistant, the first we ever saw on TV commercials or billboards, and it still has better brand recognition among U.S. consumers than Alexa OR Google Assistant. But if you first tried Siri years ago, you probably grew frustrated and stopped, either due to unrealistic expectations set by Apple or Siri’s limited functionality.

Sure, Siri may be on hundreds and millions of devices, but some people have ignored the assistant for years, and there’s no promise HomePod will change that, especially when we know price is the most important factor for most people in the market for a smart speaker.

So is Apple’s Siri strategy brilliant? Not yet, but it could be if new and existing categories focus on the user’s needs, on insights gained from its operating system, and on its low error rate. Siri just recently learned how to help you pay your bills, and if Apple chooses to go deep into the verticals it has chosen — health, productivity, and personal finance, for starters — Siri could win the day, or at least compete in a field now occupied by global giants like Amazon, Microsoft, and Baidu.