How to be effective at Google

Daz Wilkin
Google Cloud - Community
10 min readMay 11, 2018

An unstated tenet of these stories is to democratize learnings from one customer to many. Often these posts begin “A customer wanted to do [something]…”.

Today, a friend who’s the CEO of a Google Cloud Platform software partner and I walked our dogs together and I mentioned some principles of effectiveness that I’ve evolved working at both Microsoft and now Google. I’m changing roles at Google and these principles from the basis of a well-rated class that my former team gives to folks new to Google (aka “Nooglers”). The goal of the class is to provide concrete guidelines (and thus permission) to help them be effective.

These practices *aren’t* limited in their utility to people who work at Google. Nor are they limited to folks who look after customers. These guidelines are broadly applicable to help building trust and being predictable in all our relationships.

So…

I was walking with a CEO and our dogs today and he asked me to share some best practices in my experience of being effective at Google and in working well with customers. Here’s my reply:

1. Don’t Panic!

You do not now — nor will you ever — know all the answers.

When an engineer is first presented to their customer, there’s a rush of anxiety at the prospect of being asked a question and not knowing the answer. It *will* happen. It’s therefore unavoidable and you should try not to worry about it.

After 5 years looking after Google’s customers, this happened to me all the time! But, it isn’t a problem.

It’s not a problem because (a) almost all of the time, the customer’s engineers don’t expect the answer immediately and they certainly don’t expect you to have all the answers. What the customer reasonably expects is that, when you’re asked a question, that you will eventually (see note below) provide them with a (correct) answer.

NB My rule-of-thumb is that, for customers, it’s reasonable for them to expect a response to their questions within 48 hours. Customarily, I expect Googlers to respond to my customer-related questions within 24 hours (!) so this gives me sufficient room to chase with impunity (see below).

It’s not a problem because (b) a powerful way to success is to build trust-based relationships with your teammates and, in our case, with Google’s Product and Engineering folks — because someone will have the answer for you and good people always want to help you [help your customers].

Bonus points for building a sufficiently trusting relationship with your customers that, when they ask you a question, they rarely chase you because you’ve taught them to know that you always reply to their questions promptly.

Reliability is often-maligned (“he’s so predictable”) but it’s an excellent way to build trustworthiness.

2. Be Merciless!

It’s an unfortunate fact that I look like Ming-The-Merciless :-(

The intent with this guidance is not to be evil but to be disciplined in how you apply your time and energy. It’s hackneyed but true that there’s never enough time. So, our goal is to try and optimize how we use it.

Let’s start with time.

5 days in a working week (your mileage may vary; if you work fewer or more days, please use your value) and the 1950s value of 8-hours per day yields our 40-hour working week.

20% of your time for personal development.

I’m not kidding.

If you’re not learning new stuff, you’re falling behind.

Deep Work: Block one-half-day every week from now and forever on your calendar. Preferably at the same time every week. Possibly Friday afternoons because those tend to be the quietest time of the week. Be consistent (and predictable). Decline meetings during this time. Don’t do email during this time. Don’t get distracted during this time. Put on your headphones and some white noise or music. Learn something. Feel improved afterwards. End the week on a high note. When people try to interrupt or block this time, tell them “I’m sorry, [Friday afternoons] are for personal development”. I assure you, people will respect this and leave you be but you must be consistent.

20% of your time for unavoidable ‘overhead’

This is probably a low estimate but it includes email, water-cooler chats, snacks, the latest movie, those recurring meetings that you attend and wish you need not etc.

So, we’re down to 60% of the week before we’ve done any real work.

Personally, I’ll argue until heat-death of the Universe that this leaves me capacity to help 3 customers (customer projects). This is an intuitive sense that more customers requires too many context switches and because of the numbers below:

20% of your time means 8 hours/week for each of your customers

Wow! That’s (not) a lot! Let’s just consider one, hour-long meeting. Hour’s up and we’re done, right? Not so fast. What about the action items? I think — conservatively — a one hour customer meeting is between 2–3 and hours (inclusive) of new work. We’re rapidly running out time :-(

NB At Google there are two *excellent* practices for meetings. The first is that “hour-long” meetings are scheduled for 50 minutes (30-minute meetings are scheduled for 25 minutes) to enable folks who can’t transport, to move between meetings. The second is that someone always takes meeting notes.

Bonus points for sharing the meeting notes with your customers too so that everyone has a singular record of events.

Bonus points for providing an agenda (and if appropriate) briefing materials for your meetings *and* if your colleagues (!) haven’t read this, feel free to encourage them to read them *before* asking customers repetitious questions.

Let’s move on to email.

Disable Notifications: in your browser and on your phone for most everything. Unless you’re definitely interested in hundreds of micro-disruptions every day, these notifications are just irritants and bug your neighbors. Turn them off and tell me your life’s not markedly better ;-)

In *most* cases, IM, IRC/Slack etc. is the sender placing their needs above yours and their ping interrupts you. This is usually a bad thing. If that person is your customer, your manager or a significant other, fine. If the IM isn’t life-threatening, try to encourage the sender to email you instead. If you IM someone, I think it’s a good practice to include the reason for your ping in your message “… The customer’s issue (ref#) has worsened, please help me get someone from Engineering to help?”

Purists will hate me but your email inbox is an excellent way to manage your work. One of my colleagues (successfully!) manages an inbox with more than one-hundred-thousands emails in it. I’m not like that. When my inbox extends beyond a single screen, I become anxious.

Treat emails like hot potatoes.

Focus | File | Forward | Forget

Focus: If you’re going to read an email, read it. Don’t skim read it then abandon it. Don’t do this multiple times. That’s wasted energy and time. Try to get that email out of your inbox (and possibly into someone else’s ;-)) quickly.

Forward: (or reply etc.) but move it on to the next (reasonable) hop. The person (or people) who should read it or action it next. Please always review the recipients and revise these to refine the email’s audience (prefer reducing the audience over expanding). Please revise the subject line if appropriate though this will be “conversations”.

File: Yes, it’s possible that it makes you feel better to apply multiple, relevant labels but — honestly — Gmail search is *so* good that you can generally find any email easily. Label if you must. Archive please. Out of inbox is out of mind and the less you keep in your mind, the better, honestly.

Forget: For the vast majority of email the quicker you can get the email to this point, the better. For everyone.

Lists

My friends will laugh as they read this because I’m an avid list-maker. I read that our memory is similar to computer memory (DRAM) in that it requires periodic refreshes to retain information. For human memory this means, constant repetition “Don’t forget to do something”.

Which, of course, most of us then do forget.

The solution is to write to-dos down. Spreadsheets are excellent for this. List apps (e.g. Google Keep) if you prefer. A bonus from writing things down — apart from being much less likely to forget things — is that you can then remember just one thing — it’s on a list — and save yourself a bunch of cycles (and anxiety) constant refreshing your lists.

Do email in periodic stints.

Don’t open your inbox at the start of your day.

Try to spend an hour doing email and then close your inbox and do some work for a few hours.

There will always be more email.

The more email you do, the more email you’ll receive.

To misquote The Sage of Omaha, when others are telling you to subscribe to email groups, do not. More email groups means more email which requires more iterations of the above. My rule-of-thumb is to ask myself “Can I live without subscribing to this group?”. Generally, the answer is “Yes”.

3. Train Thy Customer

Controversial? Possibly. Mutually beneficial? Definitely.

A recurring theme in this post is in being consistent. In my experience, inconsistency tends to create problems.

A pattern we’ve seen several times with Nooglers and customers is that the engineer wants to demonstrate value to justify their existence, show value etc. Both of these are laudable albeit misguided aims. The issue is in not playing the so-called “long game” or in missing the wood for the trees.

In our team’s case, our role is primarily to make our customers successful by helping them get their projects into production quickly *and* removing “big boulders”.

So, the correct approach is to determine (or ask) what is the fewest number of things that I can do to have the most significant impact on this customer’s success. And only do those.

During the periodic customer reviews, removal of the big boulders are the memorable behaviors that make customers most happy (and these make our managers most happy too). The minutiae that could occupy our time are (rightly) forgotten during these reviews: no number of “But I did X”, “But I did Y” overcomes the “But we’re not in production” rebuff.

One other behavior that is counter-intuitive but productive is that, as customer success engineers, we are often able to define our role and define what we do and what we do not do. This causes some people challenges because it entails saying “no” but I encourage you to reconsider your relationship with this powerful word.

The alternative employed by many people is to demur, to not respond, to prevaricate. Is this better? I think it is not.

“No” shouldn’t be (considered to be) rude. It is clear and it is decisive.

Please read the excellent “How to get better at saying ‘No’” (link).

One important advantage in defining our role with our customer is that it allows for an honest, predictable expression of how we can best help the customer achieve success. When defining success with a customer “We need X in order to get into production”. It becomes straightforward to derive and define the path needed to get to X. And, by implication those activities and behaviors that won’t contribute to it.

As is hopefully clear throughout this story, be clear and consistent from the get-go. This reduces confusion, facilitates collaboration and promotes success. “Daz does X. Daz doesn’t do Y” says my customer “Chris does Y”. Simple!

4. Cat’s Schroëdinger (sic.)

Be decisive!

I’ve covered much of this topic above.

In my experience predictability|consistency, decisiveness strongly encourage success but — aren’t necessarily — the most obvious, easiest, least-work, least-effort path.

The upside is that, all of the behaviors described above can be incorporate piecemeal and gradually into your approach. You can become your own A|B test and see which behaviors work and which may not (I’d like to hear about these please).

In conclusion, some miscellaneous best practices.

Praise (at least) twice, criticize once. Criticism isn’t bad. This isn’t to discourage criticism. What this is to encourage, is to remember to praise people more often. It is really wonderful to take the (small amount of time) that’s needed to send some praise for work well done and make someone feel great about themselves. The World would be better if we all did this more often.

Follow-up with anyone who provides you guidance or help. Related to the above it’s easy to forget to pay it back to the folks who help us. But, it’s very little effort to message someone to tell them that their help was useful and to tell them how their help helped (you, a customer etc.).

If you’re hosting a meeting, it’s a really good practice to define an agenda even if it’s only introductions, summary of the problem by [customer], discussion [w/ engineering], review action items, next steps. This can minimally define the meeting’s structure and help you keep the meeting on track. But it shows forethought and consideration of your attendees. As above, always take meeting notes. If — like me — you’ve a terrible memory, meeting notes are an excellent way to be more efficient “Two months ago, in the meeting with …, we discussed …, you’ll be pleased to know, this is now available”.

Following the principle of minimal effort for maximum bang, invite the fewest number of people to your meetings. Be clear with optional attendees as to their expectation and or reason for inclusion (“I know you’re interested in this topic, you’re welcome to attend but not expected to discuss X”).

Start meetings, 3 minutes after their scheduled start. Even if we’re providing folks time to get to their next meeting, the folks organizing your attendees’ prior meetings may not have done so. We also want to be consistent and reward punctuality (not tardiness).

“Accept” or “Decline” meeting requests. Do or do not. There is “maybe” but it’s not very helpful and requires you to be chased with “Will you attend?”. If you’re invited and not optional, you’re needed (and you may need to decline).

That’s all!

Feedback always welcome.

--

--