Ensuring Company Alignment To Speed Up Delivery

Andrew Mullaney at UXDX Community: Dublin. Video: https://youtu.be/krH1kuORrmk

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

Newswhip and its products

[00:00:00] I'm Andrew, I'm the CTO and co-founder of Newswhip, and I'm here today to talk to you about launching. We have launched — I'm trying to count how many times we've launched — we have four active products at the moment. We've probably launched at least 20 times over the history of the company.

[00:00:27] As they mentioned, we monitor where people spend attention online and provide that to news companies, brands and agencies. We do that through four products. One is Spike, it's a real-time content dashboard. One is Analytics, which is a similar viewpoint, a bit like Google Analytics for content. API, and we have a product called Syndication which does ROI for newsrooms. This is an image that I found through Spike.

Start by asking why

[00:00:57] I want to start off with, when you're looking at a launch, I'm going to roll it back a little bit. I'm going to talk about a launch from the point of view of a product launch where you're trying to really market your product. Obviously the work is done way before that. I'm going to start with asking yourself why. Why are you developing this product, and what are the metrics?

[00:01:22] Here's a picture of me a couple of weeks after we launched our first product. I think I'm going to take a look at the product.[?] That might have done something to my joke. Technology and humor don't mix together too well.

[00:01:58] Anyway, you want to start by asking yourself why. Why are you building this in relation to the organization? Because from that you can maybe pull out some — is it a commercial product? I assume most of them are. What are your KPIs? You can kind of work backwards from there. If you're B2B or B2C, it makes a huge difference. B2B, where we are now, is much easier, I think, to understand. B2C is much more complicated and a bit of a gamble, to be honest. But we were in both, and I'll tell you some stories from B2C.

[00:02:37] From that, you can try and divide your sales metrics. How much money are you expecting to get from your product? And from that you should be able to figure out your user metrics. How many people do you need in order to be activated, in order to pay, to hit your sales metrics? From there you should be able to figure out your marketing metrics. How many people do you need in the pipeline in order to become users, in order to become customers?

[00:03:00] Another reason why it's good to talk about why is that you need to craft a story. I can't say how many products go out every day. Digital products, there's probably thousands, hundreds of thousands, but there's a lot of them. There's a lot of noise, and most people don't really care about your product. You may have worked at it for years, but they really couldn't care less.

[00:03:22] But they do like stories, and this is something we learned by using our technology, but also trying to get it out there. You'll always hear, you need to use this product because it's boom, and someone will give you a one-liner, and that's it. You still haven't figured out all the features or anything like that. It's the story that travels now, and then the product will stitch to the story. So you want to start planning that story with your product, and probably your company, from the start.

Planning the launch

[00:03:48] Obviously that requires a lot of planning. I put this slide here because most things don't go to plan, and that's how it feels like when you go to... I was wondering how I would take everyone to do that. You're fairly sharp for the time of day. Most things don't go to plan, that's all right, but it's good to start with a plan and iterate from there.

[00:04:13] Obviously the first thing you want to look at is your effort, how much time, just even estimated. I know we probably talked a lot today about story points and so forth. It is a guessing game, but it's better to guess than have no guess at all. From that you should maybe be able to figure out your finance, the costing, and the timeline as well, if you know how long it's going to take.

[00:04:35] Again, you're back to your marketing efforts. When are you going to start rolling them? When are you going to start crafting your stories? What type of things are you looking for? Content and press. All these things need to be aligned for when you're moving a product. This is something that product managers or engineers or designers can often forget as well. But if you don't launch the product correctly, there's no point in building it. So you need to be thinking about these things while you're developing your product.

[00:05:00] Same with conferences. Where are you going to find your customers? Is there a place where there's power users, specific customers? Is there somewhere you can get a high concentration of potential users? Obviously paid marketing efforts, anything from many different ones, then online pages, you probably know, like search or whatever. They can be quite difficult. Even to advertising on things like radio stations or television. I did that in a previous startup. I've seen not much effect, so I wouldn't advise it.

[00:05:31] Or partners. This is another one that people start thinking about maybe often too late. Who can you partner with to get to your users? Who is synergistic to your product, that may want to promote you because it will be good for their customers, but also good for you as well? Obviously partnership discussions take time. So if you're going to launch with somebody, you need to be talking with them very, very early, as early as you're probably doing your estimations, because contracts and so forth, that commercial stuff can take quite some time.

The product pipeline, and a note on no-ops

[00:06:07] Once you have this, you kind of should have a little bit of an idea of the pipeline of the product. Obviously it's going to start with a development phase of sorts. I know this looks a bit like waterfall, but I'm talking more from the business sense here. You're going to start with generally development and testing. You're probably going to start with an alpha then with your customers.

[00:06:27] Which is important actually, because you can get some early adopters and they will be key for your launch. If you can get alpha customers who are very happy to give you quotes, who will then go and promote the product with you when you actually launch it, then it's much better for more customers coming after the launch to be seeing people who are like them, the same use case, saying that your product is good for their use case. Obviously then you're looking at deployment, and then finally when you do launch, some scale.

[00:06:58] Actually, guys, before we're talking about — this is slightly off topic, but DevOps and no-ops. We've always been a no-ops company. There's a big debate in DevOps about whether you have ops or not. It certainly works for a while, after a while it can become problematic. I think maybe one of the issues with having no staging, which we also have no staging, is the scale side of it. You can never really test that. Anyway, sorry, I digress.

A B2C relaunch: what went viral and what didn't

[00:07:28] Okay, so I'll tell you a little bit about the story. This was a good bit back in Newswhip's history. We haven't always been a professional services technology company. Prior to this, we used the same technology to go for news consumers and tell them what the trending content on the internet was. The model for that was advertising, which as you know is difficult. Another one I probably wouldn't recommend.

[00:07:55] But anyway, at one point we were looking to rebrand and we wanted to do a new launch. The reason why I take this example from so far back is because it's a B2C launch, which is probably a bit more like a traditional launch, whereas a B2B product you can embed into your customers slower and add them on slower. It's better to do it that way, because they're going to be paid customers, so you take it slow with them. With B2C you need big numbers, so you need to figure out how to get thousands, hundreds of thousands, millions, tens of millions of people into the product.

[00:08:31] That's what we were trying to do about five years ago. We rebranded. We started with why, we rebranded. Before that we were a bit of, what it says on the tin, we used to say Newswhip, what's trending now. We decided we were moving to people-powered news. On the back of that we had a complete redesign, and we had a launch which we had been planning for a number of months. It was actually quite successful, though not in the way we would have expected.

[00:09:02] So the real lesson with this one, and I'll tell you the story behind it now, is that you have to try many different things in order to figure out what will resonate. We tried two main things. One was, as you see there, those newspapers. We got, on any given day, a picture of the actual front page of a newspaper, and then you can see in the top corner there, we redesigned one that was people-powered. So based on our analytics, we said this is what the front page would have looked like had it been powered by the people. So it very much encompassed what we were trying to stand for.

[00:09:39] Then we released them on Twitter in little bite-sized images and tagged in the editors of the newspapers. And it just went mega-viral. I have to say I didn't really expect it to, I thought they'd kind of dismiss it. I don't know how much we got, maybe 750 reshares. We were able to do it for all the newspapers. Then all the other newspapers started coming in to us and asking us to do theirs. It was just a thing for about two weeks. So it was great.

[00:10:10] We tried to create a viral video. That's what this other image is here, in the bottom right corner. This was kind of a joke. If you look in the small corner here, we created a prank, basically, where we created what was called a people-powered taxi. We kind of duped people into thinking that they'd gotten a free taxi for the day, and when they got in, a whole lot of people ran to the back of the taxi and started pushing it down the road. Then we recorded it all and released it.

[00:10:42] So the viral video didn't go viral. It just didn't. I'm not sure why. It can be a bit of a potluck. It was pretty funny, but it just didn't get picked up. And the other thing, which really described what our business did and was probably a bit more boring and not necessarily internet virality, went extremely viral. So I guess that's just a story from experience. You never really know what's going to stick. But if you kind of have a general why and theme for what you're trying to do, you can play a couple of different angles off it.

De-risking the launch

[00:11:17] Something else as well that I'd like to talk about, certainly B2B. I don't know what everyone else is. People here, are you involved with B2B products or B2C products? B2B? Anyone B2C? B2C? Crazy. I'm only joking.

[00:11:34] We always talk about, can you de-risk it as much as possible? It's hard in B2C. You really don't know what's going to work until you've got those numbers. In B2B it's a lot easier to de-risk. Something you can try and do is sell your roadmap, sell your technology, your product, your features in advance. You can get them to pay in advance. That's de-risking your uncertainty from an engineering perspective.

[00:12:01] Do your marketing on your friends. Say to your friends, what do you think, is it funny to do the people-powered taxi thing? If they say it's not funny, then it's probably not funny. You can speak to journalists in advance. You're drafting a story, you can say, look, can I talk to you off the record? This is how we're going to present our product. Is that something you'd be interested in covering, or would you be interested in getting a quote from us? Would you be interested in an exclusive, all these types of things? You can try and loop them in early.

[00:12:32] If they're not answering a call and they're not getting back to you at all, that's feedback. Maybe they've got too many people who are trying to pitch their products. Maybe you have to try and get another way into them, or craft a different story. These guys are looking for stuff that's entertaining and engaging, so I'd always say try and make it fun. Don't make it about the specifics, our product does X, Y and Z, which beats competitors' products A, B and C. Nobody wants to hear about that. You don't buy the newspaper to read that.

Premortems and deadlines

[00:12:58] Do a premortem. Has anyone heard of a premortem? Hands? Hands? Anyone here? Only one? Wow. Has anyone heard of a postmortem? It's probably better to do a premortem instead of a postmortem. The idea of a postmortem is that when something goes wrong you figure out what went wrong. The idea of a premortem is you try and break it before it goes wrong.

[00:13:26] So you work through what could go wrong here. What have we assumed in our launch, in our product, in anything, that may not actually happen? So that's about challenging your assumptions. Play devil's advocate. Be that difficult journalist. Be that time sensitive product user. Try and get that in the worst case, to really understand, do you pass your own road tests?

[00:13:53] I put this one here as well. I say beware of deadlines. Obviously you don't want things to drag on forever. But you start it by trying to achieve something, a high level why, and then if you turn around and say we have to do it by this date, and you start compromising, you're cutting corners, you're really taking risks. Really your why becomes hitting the deadline, and your why is not going to get achieved in that regard.

[00:14:18] If you're not making the progress at the pace you're expected to, then obviously you're trying to iterate to improve that. But don't keep the deadline in place and force your product out, and force your marketing out, when it isn't fully thought through. Because you'll only get one chance to launch, when you say this to the journalists and stuff like that.

[00:14:37] I've heard so many people have launched, you get a little bit of buzz, you get a tiny little bump, boom, and it's gone. And a lot of people do, to be honest with you, overhype launches. It doesn't really happen where you just flip a switch and everything goes crazy. But you might as well try and get the most out of a launch. It's something you can use to your advantage.

[00:14:59] Another example is, for some reason the press are very keen to publish fundraising rounds. It's something that we've always seen as a precursor[?] to a whole lot of worldwide press. I think our last two or three rounds, we announced probably four or five months after we actually did the round. So why tell them until we want to tell them? We tell them when we need the press. We don't tell them when we got the money, because usually when you get the money you haven't actually built anything yet. So you go spend a few bob[?] on some product iterations, get something ready, get a pipeline up and running, then tell the press you've raised money. Who cares about the actual noise that you make? All that matters is that you move those KPIs and you work towards the direction that you're trying to solve.

Launch day

[00:15:45] So, launch day. It looks a bit like this for me. Everybody's stressing out. But really I think it shouldn't be stressful at all. Actually most of the time we launch things it's usually surprisingly quiet. People are kind of a little buzzed up and everyone's excited. We've never had a big disaster or anything like that, thankfully.

[00:16:09] The technical switches are a bit different. I didn't speak too much about that, but I'm obviously in the tech side myself. Sometimes when you're moving a massive database or something like that, those ones are a bit more stressful. There's nothing you can actually do. There's a point where you just have to flip the switch and hope for the rest.

[00:16:27] You should definitely have a plan on the day, and have it communicated. The amount of times — the first couple of times we just didn't communicate things properly. What time are we going to let the press release out? What time are we going to let our marketing folks out? What time is it planned out? And try and figure out, if anything goes wrong, what will we do? What happens if the product goes down 15 minutes after the press release? Some BBC happens[?]. And that happens, I've seen that happen, I've seen that happen to companies. They've spent all this time trying to get into the biggest coverage, and then the next thing their site trips up 15 minutes later, which is heartbreaking.

[00:17:06] So expect the unexpected. Have a rollback plan. Be prepared for something that may go wrong, and that will make it less stressful as well, because if you get hit by something unexpected and you really haven't thought about it, people are going to be very concerned.

[00:17:21] Expect the expected[?]. It could go better than you expected. I'm going to tell you a story here. This is a picture of me on BBC. Why this is relevant is because I probably looked a little tired there. I had actually just finished a stag weekend. I'd flown back to Dublin and I wasn't supposed to be on TV the next day, going out to 75 million people on BBC, until the CEO came down to me and said, "I'm going to be sick and I can't do the interview. Do you mind flying to London? I'm doing it tomorrow." Obviously I had to say yes, you're never going to say no to something like that. And I was grand in the end. But you could end up, it could end up going out better than you expected. So be prepared for a phone call from a big journalist, or an investor or something. Anything can happen on these days.

After the launch: the real work

[00:18:21] Speaking of hangovers, what actually happens then after you launch the product: everyone kind of gets a little bump, boom, we get a whole lot of coverage, you see some metrics everywhere, things are looking great, and then two or three days later it kind of goes back to normal. That's when the real work starts. You've launched it, and now you've got to really start looking at your KPIs hard, and tying them back to the why.

[00:18:43] Does this KPI measure what we're actually trying to achieve? Because if it doesn't, then you have to change the KPI. But you really need to start looking very closely at it. Now you have users in the product, real users in the product, not beta testers, not friends, nothing, random people who heard about you on a news website, clicked on a link and tried it out. This is the real test of whether or not you're actually getting something working.

[00:19:10] So obviously you want feedback, iteration and improvement from there. From a technology perspective, this is when you start worrying about stability. Well, you would have started worrying about it way earlier, but now you've got real customers on the thing, and if it goes down you're going to have an impact on your numbers, because it's not available, and that's obviously not somewhere you want to be.

[00:19:37] Something like this is more of a medium to long term game. The very minute you've launched your product, all your competitors can see it now, and they will start copying you. So the race is on. They know how you're doing things potentially, or they can certainly see your product, try it out. You've written press releases, you've talked about how you built your product. You've got a lot of information out there now about what you do. And if it's successful and it's making money, then people want to compete with you. That's just natural.

[00:20:03] So you've got to stay ahead of the game, and you've got to motivate your teams. Because the launch is kind of like a party, and then you have the hangover, and then it's like, right guys, now we've got to actually put a foot on the pedal, make sure that we keep this product ahead of the game, keep reiterating it, keep going. And that's a much longer term thing than the actual launch itself. So how do you motivate more than just a launch?

[00:20:34] I think if I was to advise anything, I would say I would lower the value of the launch, I would see it as more of a marketing thing. I would see the product teams, product and engineering teams, looking at it as a much longer dev, test, alpha, beta, iterate, iterate, iterate, iterate. And most of the products out, you're just kind of in an iterating situation for an indefinite period, until you maybe do some sort of rebrand for the product. That's just a big iteration. So that's me, guys. If there's any questions, we have time to take them.