2018, Chapter 13:
Closing the Door
“Closing the door?” you say. “That doesn’t sound good! Is Schiit closing its doors? Was the year really that bad?”
In short, no and no. On the face of it, it’s been a very good year.
But, if you had to rate this year on a 1-10 scale, I’d give it about a 5.
Why? It has been an
exhausting year. The combination of parts shortages, dragged-out development, late products, R&D cul-de-sacs and black holes, process changes, test and procedural changes, and long-lead hyper-detailed planning has taken its toll.
And that’s where the “closing the door” metaphor came from. As in “closing the door on this year.” I’m ready to close the door, and look forward to 2019. I’m ready for a rest. And that’s one of the things I’ve been doing this month. That’s why you haven’t seen a lot of me around lately. But I thought I’d pop back in and wrap up the year. However, instead of three grandiose chapters like 2017, I’m going to keep this to a single chapter. You don’t need the icing and the fluff. You just want to know what we did wrong, what we did right, and where we’re going.
So without further blather, let’s get into what we screwed up in 2018:
What We Did Wrong
When I closed the 2017 edition of Schiit Happened, I was pretty damned upbeat about our product plans for 2018. Hell, I even said, “Barring any major complications, you’ll see Sol (the turntable) and a Gadget (retuning technology) in 2018.
(insert a single lonely cricket chirping here)
Yeah. Because you got neither a turntable or a re-tuner. You also got a couple less products than I thought you’d get.
I could be disingenuous and hide behind the “global parts shortage” excuse, or an even bigger jester and say, “the tariff situation is causing disruption in our supply chain,” but both of those are pretty much entirely bullschiit. The global parts shortage thing has been a pain (I’ll get to that), but the tariffs haven’t really hit us (much).
No, the reason you didn’t get a Sol was simple:
it’s complicated. It’s our most complicated mechanical product, ever. We are rightfully paranoid. That, combined with Mike and I both being busy with internal challenges, meant that progress on Sol was slower than expected.
“Buutttt…I want a Sol!” you cry. “Are you gonna tell is it’ll be out in 2019 and miss again?”
Possibly. But that would be fairly improbable, because Sol is in a much different place now. More on that later, when I get to Where We’re Going.
The reason you didn’t get a Gadget is also simple
: it’s complicated.It’s our most processing-intense product, and Mike and Dave and I are still debating the best way to package it, and what its feature set should be. Because we don’t want to create products just for the sake of creating products; we want to create something that you’ll really enjoy (and, to be frank, something that doesn’t need to be upgraded pretty much immediately because we figured out something else that it could do.)
“Buttttttt….I want a Gadget!” someone else cries. “Are you gonna miss that one too?”
Hmm, I’m less sure about that one. Ask Mike and his team. The bottom line might be: when it makes sense (and its feature set is complete), then yes, you may be seeing something from us. If not, not.
The reason you didn’t get a couple extra products I had planned is also simple:
it’s complicated.In that the products turned out to be more complex than I figured. One, Aegir, should have been a cakewalk. It turned into a very complex engineering challenge. Another went through a couple of take-it-to-the-end false starts before I decided to really, really make it right.
And everything was complicated by what I’ll call…
Global Stupid #1: Too Much Tied Up In Our Own Shorts
Both Mike and I spent way too much time in 2018 tied up in our own shorts. Navel-gazing into the engineering abyss, wondering why things weren’t working exactly right. Running into cul-de-sacs and dead ends and shaking our heads, unsure of how we got there. That was engineering in 2018.
Examples?
Mike had the USB project, which was supposed to be a short, maybe 2-month runup to a working prototype…and that ended up eating probably 3x that amount of time. I had a new topology that I’m calling Nexus, which worked better than I expected on first turn-on…until I tested it a bit more…until I dug into what happened with imperfect parts…until I dug into thermal properties…and my 2-month dev turned into, well, pretty much all year.
“So why is this bad?” you might ask. “You guys need to develop new stuff, and sometimes it doesn’t go as well as you expect. Big deal.”
Agreed. Not a big deal. That is, until both of us are worrying about projects that might be several generations out.
When that worry eats too much of our time, we can miss things. Things like updates to current products that should really be considered. Things like handing off Sol development to another team member so it can move along faster. Things like ensuring that soon-to-be-released products are really, really done.
Things that turn into things like:
Global Stupid #2: Few Intros, Late Products
Let’s cut to the chase: we didn’t introduce enough new products last year. In fact, we introduced only two really new products, and one of them was late.
- Lyr 3: this one wasn’t late. This one actually worked pretty well. But it is only one mid-range product release, which is probably too little to spread out over the course of a year. And by “probably” I mean certainly.
- Modi 3: this one was a LOT late. I really wanted this one to happen in May, following Lyr 3 by a couple of months, and setting us up for a summer where we could introduce Sol, and then maybe Aegir, and then maybe a Gadget, and then maybe something else, and that would be the year. This one was WAAAYYYYY late largely due to the global parts shortage. There were a couple of parts that really bit us in the butt. We also spent more time than anticipated battling backorders, which didn’t help.
Now, if you’re counting, you’ll also note that we introduced the Octal LISST, the Universal Multibit card, the Universal 4490 card, and the Yggdrasil Analog 2. All LISST are now dead, victims of relatively slow sales and difficult assembly. Sorry about that. We need to stay focused. Don’t be surprised if we get even more focused next year?
The upgrades? The Universal Multibit and Universal 4490 cards went quite painlessly. We’ll talk about those later. In the meantime, let’s talk about…
Global Stupid #3: An Upgrade Debacle
The Yggdrasil Analog 2 upgrade was the first thing we announced last year, and, in retrospect, it was not the brightest thing to do.
Why? Lots of reasons:
- We were less than a year out from the introduction of the Gen 5 USB, so it was entirely possible that someone had recently sent their Yggy to us for upgrade. Sending in your DAC twice in one year for hardware upgrades tends not to make people happy. (And yes, we totally understand this. The timing was horrible. We won’t be doing this again.)
- The Analog 2 Upgrade REQUIRES you send the Yggdrasil back to us for firmware changes. This is not great, even with a queueing system. (Again, we understand. We hear you. We will make it better.)
- The Analog 2 Upgrade uses the same DACs as the old Yggdrasil analog cards. This makes the upgrade pretty hard to explain. It also makes it very costly, because there are four $64 DACs on the analog boards, which cannot be recaptured and resold as new. So it’s a very costly upgrade.
So, with one upgrade, we had a chance to irritate and inconvenience a whole lot of people. Which we took. Not our brightest hour. Again, going forward, we will be careful with timing of future hardware upgrades and we will be working to ensure that the maximum number of upgrades do not have to come back to us (or a distributor) to be performed.
Global Stupid #4: Still Too Much Backorder
Worst of all, the Analog 2 Upgrade didn’t help our backorder situation. With complex assemblies, long lead parts, and a large upgrade queue, this isn’t surprising. The problem was, we really wanted to eliminate backorders in 2018, and fill all of our channels (like distributors and Amazon.) Keyword: wanted.
A quick look at this forum over the past year shows how successful we were at eliminating backorders, which is to say: not very. Not until very recently, that is. We finally are able to fill distributor and Amazon channels…just about 4 months later than planned. And we’re still currently out of stock on Valhalla 2.
One of the things that really hurt backorder, even beyond the upgrade queueing, was the global parts shortages. This made kits slip one, two, or even three months. And when you’re looking at a schedule that’s held up by a couple of $0.20 parts that you simply can’t find anywhere, it can get very, very frustrating.
So, despite our best intentions, our plans didn’t pan out. Backorders continued. Perhaps we can do better next year. That’s certainly our intention, anyway.
We’ll see.
Global Stupid #5: Customer Service Fix #2 Took Too Long
Late last year, I changed customer service over to a ticketing system, due to emails being missed and too much miscommunication. This seemed to work very well—it gave us the visibility we needed to ensure customers were getting the support they needed, and it allowed us to assign the more complex questions to a higher level of technician.
But this only works when everyone is responsive. And when one of our techs drops off the map and doesn’t tell us (and we didn’t catch it—that really is our fault), then the same old problems came back—missed emails, no responses, and a level of support far below what we should be shooting for.
I tried to make it work with our technician, and I tried to shore up, but, let’s be real: we let it sit too long. Coupled with an ordering staff who had their all-too-human moments when confronted with a customer coming in hot, we really needed to make some changes.
So, very recently, we did. We now have a different tech support staff, additional oversight at a higher level, and ordering staff have been retrained, with new policies in place that allow them to be much more flexible than in the past. I’ve been watching progress closely, and things seem to be operating much more smoothly.
However, I feel compelled to add a personal aside here: everyone you interact with here at Schiit is a very smart human being, largely unconstrained by corporate trouble-trees and best practices. Being nice gets you very, very far.
What We Did Right
So is there good news from 2018? Of course. All that R&D time burned, all the process tweaks, all the production procedural changes, all the new reviews and checkpoints make us a markedly different company than we were at the end of 2017.
In short, last year was a year to build new foundations. Due to this, we’re on much better footing going forward (or at least I’ll keep telling myself this!)
I’ll keep this one short, because every time I try to pat myself on the back I hurt my arm:
Good Stuff #1: Better Controls, Better Products
Until the start of 2018, we were still operating a bit too much by the seat of our pants. Yes, we caught problems as they came in—but sometimes there was no official process by which to review new runs and have clear approval. Many, many of our products were tested with long, laborious listening tests, rather than be largely automated.
In 2018, we changed that—first by deploying Avermetrics analyzers throughout the company to automate testing of some of our most difficult products, including phono preamps and multibit DACs. At the same time, we created new processes to review new and ongoing products from chassis to firmware, so we now have much better controls in place to ensure great products.
These new controls are why the product launches this year were largely glitch-free, including the Modi 3 launch—a very high volume product.
In addition, these new controls, tests, and procedures are in place for when we introduce even more complex products, like Sol. Sol may (er, I mean, WILL), have its own learning curve, but the fact that we’ve taken these steps should make it easier. I hope.
Sounds simple? Sure, if you work for a Fortune 500 company. Remember, we’re not all that far out from the garage years. But that’s now changing…and fast.
Good Stuff #2: Reorg that Works
Sol slipped all through the year largely because we didn’t know who the hell was supposed to wrangle this very complex product. Mike was busy, I was busy, we’d get together and look at each other and say, “You know, we really need to get these last few parts qualified,” and then we’d drop it. Because neither of us had time, and neither of us was THE person in charge.
That changed when Mike had the great idea to give Tony the Sol project. Tony’s been with us, like, forever, and he’s much more focused and detail-oriented than Mike and I. He’s been managing the Sol now through the first articles and final prototype assembly. And now, finally, I can say that we really are getting there.
“Yeah, I heard that before,” you say. “What does ‘getting there’ mean?”
Here’s what it means:
- All parts have been qualified
- All parts are ordered
- Most parts are in-house
- Final cosmetic machined castings for plinth and platter are imminent
- Plinth and platters are 6-8 weeks from approving the final cosmetic parts
So, again, I really do think you’ll see a Sol in 2019.
Really.
Stop laughing.
But the reorg doesn’t stop there—and, in fact, training will be starting soon for all the changes coming in 2019. More on that next year.
Good Stuff #3: Much Deeper IP Portfolio
Mike and I spent so much time last year with our heads up our butts (er, I mean, engaged in research and development) that we now have a deeper IP portfolio to pull from, both on the digital side and analog side.
On the digital side, this includes our own USB interface (in testing now), additional Gadget functions, foundational work on a transport, and a few (yes, a few) other digital things I can’t talk about right now.
On the analog side, this includes more products with Continuity, starting with Aegir, and a commitment to Continuity being the right way forward (rather than Class A, no matter how you define Class A). It also includes Nexus, a new balanced topology that you’ll see in at least one…no, wait, two…products next year. I’ll do a chapter on that when it hits. Unfortunately, that will be fairly deep. And there are a couple of other things I can’t talk about yet either.
And yeah, I know, “still just a marketing company.” LOL.
Good Stuff #4: Killing Sacred Cows
Last year, we killed LISST. Doesn’t sound like much, right? But LISST was part of the promise of Mjolnir 2. Have a tube amp—and switch it to solid state with LISST whenever you wanted!
The problem was, most people didn’t want to convert their tube amp to solid state.
Yeah. Huge shock. Maybe we shouldn’t be so stupid in the future.
But, in any case, we introduced octal LISST for our new line of products. And sat back and watched as they didn’t sell. In the past, this would probably mean that we’d keep making them for a while to see if demand picked up, while also balancing the fact that no sane PCB assembly company wants to make such a fiddly, goop-filled device, and therefore never wanted to find time to make them, which meant we’d really be out of stock most of the time, which meant that everyone would have to answer questions about why we weren’t in stock…
…so, instead, we decided to do something smart: we killed the LISST.
Now we have more time to concentrate on products that sell. We can even ensure a good supply of another product that kills even more complexity: Modi 3. Modi 3 replaces both Modi 2 and Modi 2 Uber, making putting together a great small desktop system even simpler.
Good Stuff #5: A New Resolve
“Reorganization, killing sacred cows…holy moly, what’s next?” you ask.
Simple: upending the whole thing. Last year was, overall, boring, slow, and safe. It was a slog. It had us worrying about things we didn’t need to worry about, and missing new opportunities because we were working on the next next great thing, or worrying about minutiae that really didn’t matter.
It’s taken a while, but the energy has changed significantly in the past few months. We’re working together, faster and more effectively than ever before. We’re asking, “Why do you have to do it like that?” and throwing out old ways that have been holding us back. We’re cranking through multiple new products that are really, really shocking in one or more ways—performance, or price, or just the whole way it works. We’re making changes and variations on current products to see how we can make them better—and exploring ways to bring in these changes, transparently, and improve the whole line.
In short, we’re working at a pace, and with a spirit, that I haven’t seen since we decided to reinvent the headphone amp and affordable DAC market when we started. 2019 is going to be an incredible year.
No. Really. You’ll see.
Where We’re Going
I’d love to give you a detailed product roadmap for 2019, but you know that’s not really possible. Not only are there a lot of what-ifs and maybes and possible interruptions and maybe even more cul-de-sacs ahead, there’s also the practical reality: we’re going to be replacing, updating, and eliminating quite a few products. I can’t Osborne Effect us out of existence.
“How many products we talkin about, bub?” you ask.
Well, a lot more than last year.
“Hell, a dead dog with three broken legs could intro more products than you guys did last year, so that ain’t so impressive,” you say. “Give me a number. I won’t hold you to it.”
Yeah. Uh-huh. We’ve heard that one before. If I say, like, “six,” and then we only deliver five, then you’ll say we missed the mark. If we do seven, then we’re flooding the market and reducing the resale value of your recent purchase.
So, no number. How about a range?
And how about a range that might go from half a dozen to double digits?
“Oh no, no no no no no way, you did that before, you introduced 12 products one year and you said that was too much!” you say. “Now you think you might do it again? Whatareya, nutz?”
Maybe. But let’s tot it up:
- We really think we’ll be starting with Aegir in January or February.
- Then there’s gonna be a major intro in March.
- After that, we have Sol sometime in Q2. I hope.
- But before that, we're going to have four more intros. Yes. Four.
- Oh yeah, and then there’s a couple of big intros around summer. Or maybe 3, depending on how you're counting.
- And let’s not forget one or two nearer the end of the year.
That’s right up there at 11.
And that might not be all.
“Well, you just proved it,” you say. “You guys are a whole buncha dumbasses. You already said that was too many products, a few years back. And I thought you guys wanted to go leaner and meaner and have a simpler product line.”
Well, yeah. It’s definitely too many products…
…IF you’re talking about adding products to the line in a traditional manner…
…and IF you’re talking about the same-old, same-old way of doing things…
…but who said we were talking about either of those things?
Yeah. It’s not just product intros we’re talking about. It’s not just upgrades. It’s not just expanding the line. Beyond that, I really can’t say any more. You’ll find out what I’m talking about soon enough (well, before midyear. I hope. Oh please make it so.)
Because 2019 is a new beginning. A year to shake everything up. A year to take chances. A year to break things. Sol may radically change the complexion of the company. And at least one other product may have the same effect. Beyond new products, you’ll see at least a couple of technology intros. You also may see some familiar names go by the wayside, or be reborn in radically different ways.
Bottom line, if there’s one word for 2019, its this:
intense.
Buckle up.
