Last Updated on September 6, 2020 by Alex Birkett
As my company grows and we onboard new employees, I’m spending more time thinking about leadership, management, and giving advice to speed up others’ progress. Writing down some career advice for a new employee, I realized something obvious:
Any advice I give says much more about me than it does about the receiver of the advice.
Whether it’s hindsight bias or narrative fallacy, it’s easy to find patterns in the data points when looking backwards at your own life. We patch things together to find something coherent in the randomness.
Even if there’s truth to the patterns you’ve found in your own winding path, it’s not the case that these data points will apply equally and effectively to other people’s paths. Learning R may have benefitted me, but it may also be a waste of time for someone else depending on their goals and interests.
There are too many variables:
- Randomness and inscrutability of causal factors – how do you know the things you think were impactful were actually the things that mattered?
- Timing – the same tactics applied at different time windows can have highly variable outcomes.
- Context – this is the big one. What works for one person (or company) might not work for another, for a massive and maybe infinite amount of reasons.
It’s not that I think advice is unhelpful. In fact, I believe one of the true lessons of my career has been listening to the advice of mentors who’ve already traveled the path I want to travel. I hope my advice helps new employees and anyone who seeks it, too.
The point I’m trying to make is this: when you get advice or learn something from someone else, contextualize that advice with the specific situation from which they’re giving it.
Shouting Down the Echo Chamber
The marketing world is absolutely stuffed with debates that never seem to die. Some are, in my opinion, based solely on differences in definitions:
- What is growth marketing / growth marketing vs. traditional marketing
- Data-driven vs. data-informed
- Is the acronym ‘CRO’ really the best one for what conversion rate optimization actually is/does?
Labels and words matter, of course, but these debates typically wind on endlessly because the interlocutors are talking past each other, definitionally.
There’s another set of debates that is much more meaningful, though. These questions are things like:
- How many experiments should you run per week? Should you optimize for tempo or experiment quality?
- Quality vs quantity in content marketing — how often should you publish, how long should the article be, and should you write 10X content or expert roundup posts?
- What percentage of time should you spend on content promotion (some say 80%, some say none)?
- Should you invest in only one marketing channel or many channels simultaneously?
- Should you build your own experimentation platform or buy a third party tool?
- Is paid acquisition a valid and scalable acquisition channel by itself?
These are really great questions. The frustrating part about them is they all end with the same famous answer, at least if you’re being intellectually honest: it depends.
A Tale of Two Companies: Experimentation
Let’s look at one of these questions: how many experiments should you run per a given time period?
While there are philosophical differences that determine the different answers (some think with enough research you can craft a hypothesis that has a higher likelihood of winning, some believe we’re mostly ignorant as to the outcome of a given test and we should just test as fast as possible or the widest variety of experiences possible), the biggest difference in approach comes from this: the size of the company and the amount of traffic you have to test on.
Is it any surprise that booking.com, Microsoft, and Google executives advise you to “test everything” and “run as many tests as possible?” They have unlimited traffic!
There’s basically no opportunity cost in running an experiment, and in fact, the expected value of any given action at that size may even be negative. This means that, by running the experiment, more often than not they’re actually trying to mitigate risk and potential loss, instead of aiming for double digit percentage uplifts.
A startup doesn’t care about not losing; they want to lift their conversion rates so they can acquire more customers and hit some semblance of escape velocity in their growth.
Additionally, a startup may need to move a core engineer away from building features or fixing bugs to writing “throwaway code” on experiments that don’t end up moving the needle. In many cases, there aren’t enough resources to allocate to run the quantity of experiments you’d like to at a startup, so you need to bundle up your tests and aim for the fences.
Larger organizations are trying to protect the downside, and of course, running an A/B test on every code push is a great way to do that. They typically have experimentation platforms and internal tooling built specific to democratize the ability to test and analyze experiments. Most companies don’t have that.
Context matters. If you’re a startup, taking advice from Microsoft may lead you down the wrong path.
Why Your Startup Isn’t HubSpot and Shouldn’t Try to Be
Another example: content marketing.
The central debate in content marketing is probably “quality vs. quantity.”
There’s some definitional obscurity here, too.
What we mean by quality and quantity can differ, but what the debate is usually discussing is this: should you write a lot of content as fast as possible, knowing that quantity has a quality all to its own (stuff will start to rank naturally), or should you take a ton of time and effort to craft each individual piece of content to be as perfect as possible (it’s hard to stand out, quality is the only way to do so)?
HubSpot, and much of the ‘Content Marketing Institute school’ of content marketing, hinges on “publish as much as possible.” It’s still high quality, but it’s an article on Content Marketing Institute, HubSpot, or SAP’s blog is definitely not the same length/cost/research level as a given article on CXL, WaitButWhy, FirstRoundReview, or Moz (especially in their early days).
Of course, content marketing programs have different goals, and that can factor into your approach.
There’s a philosophical difference between, say, Databox’s approach — which, despite not knowing their team, I can safely say is to drive as much organic traffic and leads as possible through repeatable expert roundups — and CXL’s approach, which is clearly not always SEO-driven and thinks of content as a way to build a brand.
Furthermore, a brand like ButcherBox writes most of its content for its current customers, and it doesn’t even look at content as purely an acquisition channel like most of the B2B world does.
But beyond that, what accounts for the difference in approach?
Company size, and more importantly, domain authority.
Two companies with the exact same goal (drive business results through SEO-driven content) should have different approaches to the extent that their domain authority differs. HubSpot, with a Domain Rating above 90, can publish almost whatever they want and it will rank within a week. Endear, with a Domain Rating near ~30, needs to write fucking epic content every time and build a ton of links to each piece to have a shot at competing.
As your domain authority increases, your quality bar can lower, and all things equal, you can begin to write content faster without paying as much per individual article. Moral and philosophical arguments aside, the economics here are pretty clear: why pay $1k for an article when you can pay $200 if it gets the same results? Why write 1 article per week when I can write 10, if each article gives me similar results?
Quality vs quantity is a meaningless debate, not because the questions is not important, but because the answer varies based on who is answering it.
One Channel or Every Channel?
A central question in the mind of any strategist is “how do I distribute my efforts?” In the context of marketing, that questions warps into something like, “should I diversify my channels or should I dominate one channel fully?”
This breaks out into different specifics depending on your context. In e-commerce, the question is whether your business should invest in “omni-channel” marketing. In social media, should you spread yourself across all popular channels (Twitter, LinkedIn, Facebook, etc.), really own just one or two, or constantly seek the newest hot and rising platforms?
Again, in almost every case, the answer differs depending on the context (unless, of course, that context consists of breathless pundits and thought leaders shouting at you that you *need* to invest in omnichannel lest your company falls behind).
I picked up a great old (1999) advertising book called Secret Formulas of the Wizard of Ads that explained this well. As the author, Roy H. Williams, puts it, “Strategies, concepts, and technologies that are proven winners for big companies are often counterproductive for smaller ones. What’s good for the big corporate goose can be poison for the independent gander.”
He continues, giving the example of media mix in advertising:
“There’s a lot of truth in the religion of media mix, but it is a truth inappropriate for the small business owner. The idea of a media mix assumes that your advertising budget is adequate to do a good job in each part of the mix. Coke, Pepsi, Chevrolet, Procter & Gamble, and the other big boys are able to accomplish a media mix without being forced to compromise any part of it. They can mix radio, television, newspaper, magazines, and skywriting without having to do anything halfway.
Is this true of your company? Do you have this kind of budget? If not, I recommend that you do one thing well rather than two things badly. You may be able to do two things well, but I’ve never worked with a company that could do three things well. But then I’ve never worked with anyone who had more than a couple million dollars to spend on advertising.”
First Principles to the Rescue
Despite all of that variance, I think we can agree upon a few things:
- You can ignore most gurus and thought leaders (for more reasons than just the issue talked about here)
- First principles thinking gives you flexibility to apply advice and tactics contextually.
For instance, in the case of experimentation, here’s a first principle that’s hard to argue with:
Experimentation is a research methodology that helps reduce uncertainty, though never fully, in order to help us make better future decisions.
Now, from that, you can decide “how much uncertainty am I comfortable with?” and “how much am I willing to ‘spend’ to reduce X amount of uncertainty?”
Startups, typically, can live with more uncertainty, though opportunity costs will kill them. The cumulative value of small amounts of uncertainty at a large company can drastically alter the course of its products and revenue. The approach is different, but the first principle is the same.
In content, it’s a little bit trickier because beliefs diverge. At its core though we can say, “content marketing is a cost efficient but time intensive acquisition channel,” and then we can figure out what the tradeoff in cost, time, and effectiveness is at different stages.
In the beginning, that may mean paying or spending a lot of time to rank a single article, and in the latter stages it may mean writing so fast and frequently that you steal all of the valuable and valueless keywords in an industry. In one case, you’re playing the differentiation game and in the other you’re playing the scale game, but both are derived from the same rules and beliefs.
Despite All That, Here’s 43 Pieces of Marketing Advice
Okay, so the answer to most questions is “it depends” (or better, “it doesn’t matter.”). So what? What do you do with that?
Well, like I mentioned in the intro, I’ve been thinking about what I’ve learned about marketing and writing down some lessons. So instead of just letting those ideas rot in my college-ruled notebook or Notion folder, I’ll jot ’em down here. Think of this like a list of lessons written for a younger version of myself, ambitious little marketer that I was.
- Best practices are like training wheels, but first principles thinking is how you create interesting work. Chasing the median passes the exam, but doesn’t create trail blazing work. Learning to think for yourself and spending a lot of time thinking is one of the biggest investments you can make (even if it doesn’t *feel like work*, it is).
- Advice almost always says more about the giver than the receiver. Even worse, advice on the internet is generally incentivized to either be a) simplified to the point of banality and uselessness or b) inflammatory and contrarian to spark engagement. It’s all good fodder for thinking, but in the end, your best bet is thinking a lot for yourself and running a lot of experiments.
- Find mentors who can be straightforward and honest and give you constructive feedback. Instead of seeking positive feedback like most people, try to seek out the painful stuff. That’s the only way to get better.
- A good mentor is probably the most valuable “shortcut” you can take for your marketing career.
- Just as good as a mentor: get an accountability group of peers that you can learn from, preferably in different companies and/or industries. I meet monthly formally with my group, but I’ve also got informal peers like this I meet with regularly for lunch.
- Sometimes you’ll have to make a trade-off between what works and what you believe is right to do. I biochemistry already gives me trouble getting a good night’s sleep, so I try to err on the side of principles instead of the common excuse, “oh well, the data says it works!” Of course, there’s a balance, but solely seeking “what works” may lead you to build a website with blinking CTAs like a Christmas tree, copy like a used car salesmen, and an arsenal of basically plagiarized content (it’s cliche to even mention the thought leaders you know that are like this).
- No one gives real cutting edge advice or information in blog posts or business books. Everything interesting is told during the happy hour at conferences, coffee meetups, or informal settings.
- Similarly, the incentive to create case studies doesn’t align with your ability to learn anything interesting from them. Avoid most.
- You can get obsessed and spend n months (sometimes 3-6 months, sometimes years, sometimes a few weeks) diving deeply into a business topic and learn most of it, at least to the level that you’re proficient (not mastery). Works with stuff like branding, differentiation, Facebook ads, whatever. I find these things are to be learned as necessary, when problems and challenges require those skills, and the foundational skills are what matter most (communication, data analysis, negotiation, etc.). Most importantly, learning how to ask really good questions is the centerpiece, and learning how to learn (meta-learning) makes everything else easier. Good book on this is the 4 Hour Chef by Tim Ferriss or The Art of Learning by Josh Waitzkin.
- You get good by taking lots of swings. Optimize your learning rate α not through perfection and reading lots of theory, but by trying a lot and hopefully getting good feedback each time.
- Nobody counts the number of items on a listicle, so if you just say a high number it will sound impressive. Peak skyscraper method.
- That last item, of course, was ironic (who is counting these tips?), and this one is breaking the fourth wall. Turns out, you can do whatever you want in marketing, including self-aware blogging and copywriting. This is yet one approach to “voice,” but another one could be earnestness and sincerity. Another could be that sort of corporate sound jargon-heavy enterprise copy you find on bland SaaS homepages. I’d avoid that stuff.
- Copywriting is the skill with the highest expected value, at least in terms of how much cash you’ll make if you’re good at it.
- This is one of my favorite articles on career advice (ignore the title pointed at a “young man” it’s really aimed at ambitious young people, i.e. Ryan Holiday is writing to a younger version of himself).
- Actually, Peep Laja just wrote an insanely valuable “advice to my 30 year old self” thread that is worth reading. These types of “letter to my younger self” write-ups are awesome and you should eat them up. Experience is expensive.
- Get good at hard things that other people don’t want to spend the time getting good at. This changes over time. For a while it was data science/ML, but SaaS tools (e.g. Squark, DataRobot) are getting good at productizing that. I’m not sure what that thing will be in two or five years, but keep an eye on what is important but no one seems to have really figured out or mastered and own it.
- Get good at SQL. It’ll never not be useful if you work in tech.
- Get good at writing (clearly, persuasively). It’s the most underrated business skill.
More Marketing Advice on Building a Personal Brand
- Only write (or speak or do videos or whatever your medium of choice is) about what you deeply give a shit about, otherwise it’s a chore. People respond instinctively to passion and authenticity.
- Do so consistently once you get in a groove. You want a surround sound effect. Repetition + message resonance is how you win in advertising and marketing.
- Find a channel that you like using (if you hate TikTok, don’t use TikTok. If you hate blogging, don’t blog).
- That said, if you can build an email list, do that. It’s the highest value channel you can have at least on a personal brand standpoint (and probably true from a company marketing strategy standpoint or for business owners too). Email marketing is super underrated and usually done poorly, but the money’s in the list.
- Network with others doing the same thing as you (look at how podcasters all go on each others’ podcasts, writers tend to form ‘salons’ with other writers)
- Build up your own website/blog. I actually somewhat regret writing so many (good) guest posts at this point, because I just imagine how valuable that traffic would be if it was flowing regularly to alexbirkett.com and not random analytics software or marketing agency blogs.
- If you can, kill two birds with one stone. Working on your personal brand takes a lot of time and effort so try to do so in a way that benefits your current company/day job. This is obvious if you’re running your own agency or something (your brand = new leads/sales/business value), but you can do it if you work in a marketing role, especially content marketing. Most HubSpot bloggers are basically influencers because of how many people have read their content. I got paid to write about CRO and analytics when I was at CXL, and it formed the basis of much of my professional and personal network as well as my small personal brand + launching point for this blog.
- You can have a valid marketing career without being an influencer or having a personal brand. This is how most people go through their lives and career.
- I have a hypothesis that a large percentage of marketers had early artistic career ambitions and therefore fill that hole by seeking similar external validation and recognition through their marketing careers. That’s why there are so many “marketing celebrities” and why some conference talks are so theatrical (I’m only throwing rocks because I once wanted to be a rockstar and still write crappy indie/punk rock songs in my spare time).
Marketing Strategy and Management Advice
These will be more specific marketing tips (not just career growth or personal brand stuff). Will also cover some leadership advice for marketing managers.
- Tear down/delete your roadmap every six months and re-establish your backlog of ideas/plays. Good ideas will crop up again and you’ll clean out a lot of fluff that sounded nice but is unimportant.
- Quantum management: tell them where you want to go, but not how to get there.
- Unintentional advice from a former boss: “if you can’t pay them more, make them feel important.” If you think about it, that explains a large part of how the world works, and why you should, when possible, say fuck status/title and just take the money.
- Invest in one channel to grow and diversify for defense. To find your cash cow channel will require experimentation. I recommend a ‘multi-armed bandit’ approach, trying multiple channels until one feels like it clicks. When one clicks, put more effort into it and less into the others.
- That said, there are only a few truly scalable channels. These change over time, but they tend to be SEO, paid acquisition, virality, sales (and misc. like partnerships, but rarely). You can typically take a good guess or two to figure out which has the highest impact potential in your context.
- Organic channels are superior to paid because you don’t need to keep pouring gasoline on the fire, but they’re not “free traffic” and they’re not easy, fast, or optimal for all clients. Not everyone should have a blog. Even if your company will eventually have a blog, maybe it’s not the best time to focus on it now. There are always opportunity costs. What channels and efforts you invest in shouldn’t be built on the opinion of thought leaders selling books, courses, or SaaS software that helps you do the thing they’re selling.
- Copy is typically more impactful than design, so if you’re trying to prioritize efforts on a landing page or email message, start with your messaging and storytelling. That said, Andrew Anderson has said he’s found the opposite to be true in his experience. Context is everything remember. Experimentation is the way.
- I used to think webinars sucked and they didn’t work. Then I thought, “huh, I wonder what webinars would look like if they didn’t suck?” And that was a fucking epiphany. There’s no real valid way to say a tactic is good or bad, it’s all in the execution. Execution tends to be high caliber in the beginning stages of a tactic or because it’s novel and the innovators are mopping up with few competitors. Then people flood in, copy the tactics, water everything down, and people declare “X is dead.” Turns out, you can make anything awesome.
- With webinars, at CXL Institute, I thought about all the stuff I hate about webinars: 20 minute introductions of the speakers, long sales pitch at the end, high level/slide deck style presentations, beginner stuff that I’ve seen before. I scrapped all that. Didn’t intro the speaker at all or even appear on screen (people didn’t give a shit about me, and they already knew the speaker — they signed up!), webinars were screenshare only and demo’d actual skills, and we went way beyond the beginner stuff. In short, they were actually valuable. Webinar signups started at like <100 peopled and went up to 1-2k depending on the topic.
Read these books
Trust me, these ones will help you out:
- Meditations by Marcus Aurelius
- Alchemy by Rory Sutherland
- Trust Me I’m Lying by Ryan Holiday
- Antifragile by Nassim Nicholas Taleb
- Influence by Robert Cialdini
- Anything by Robert Greene
- David & Goliath (and Outliers and Tipping Point) by Malcolm Gladwell
- How Not to be Wrong by Jordan Ellenberg
I’ve also got lists specifically for CRO books and A/B testing books if you’re interested in those particular fields.
Wrap Up on Marketing Advice
Context matters more than the specific advice given. Microsoft’s approach to experimentation won’t work for your startup. HubSpot’s approach to content marketing won’t work for your small business.
That said, you can learn a lot from mentors, books, and those who have been there and done it before. You can save a lot of time and “dumb tax” by learning from others first, just use critical thinking and don’t copy + paste others’ life experience and marketing advice.