Marx Twain

The beau­ty of contradictions

We’re liv­ing in stu­pid times and it’s very hard for any­one run­ning a busi­ness to sep­a­rate truth from fan­dom, oppor­tu­ni­ty from hype, opti­mism from wish­ful think­ing, and fear-mon­ger­ing from fact-based con­cerns. We need to train our­selves to question.

Truth is usu­al­ly some­where in the mid­dle (some­thing Marx was actu­al­ly right about). 

There are two ubiq­ui­tous top­ics you’ll find in any 2025 pre­dic­tion from every tech bro, dig­i­tal prophet, fear­mon­ger or doom­say­er out there. Let’s find the (mid­dle) ground in all that hot air.

Pre­dic­tions & con­tra­dic­tions 1: AI or Die

Since Chat­G­PT entered our lives we’ve been immersed in this mantra-like obses­sion about arti­fi­cial intel­li­gence, and a future that will leave us behind unless we change, right now, by adopt­ing all of their tools, regard­less of their util­i­ty, or else. This is not about tech­nol­o­gy. It’s about mon­ey, pow­er, and grow­ing the Ponzi Scheme before investors find out. I’ll prove it to you easily. 

Accord­ing to Wikipedia, Arti­fi­cial gen­er­al intel­li­gence (AGI) is a type of arti­fi­cial intel­li­gence (AI) that match­es or sur­pass­es human cog­ni­tive capa­bil­i­ties across a wide range of cog­ni­tive tasks. That’s an out­stand­ing achieve­ment. Almost god-like, if you want to delve into the philo­soph­i­cal ram­i­fi­ca­tions (I won’t).

Do Microsoft or Ope­nAI share that def­i­n­i­tion? That ambi­tious, piv­otal mile­stone in human history?

The two com­pa­nies report­ed­ly signed an agree­ment last year stat­ing Ope­nAI has only achieved AGI when it devel­ops AI sys­tems that can gen­er­ate at least $100 bil­lion in prof­its. That’s far from the rig­or­ous tech­ni­cal and philo­soph­i­cal def­i­n­i­tion of AGI many expect. This year, Ope­nAI is report­ed­ly set to lose bil­lions of dol­lars, and the start­up tells investors it won’t turn a prof­it until 2029. (TechCrunch)

If they were sell­ing 100B of fid­get spin­ners, they’d call it break­through tech. It’s just mon­ey. It’s a rigged game.

It’s a race for dom­i­na­tion based on lies, loss­es, lob­by­ing and leapfrog­ging from one fool to the next. 

We’re pour­ing 500B into a com­pa­ny that’s not mak­ing mon­ey at all and call­ing it ​“growth”.

Pre­dic­tions from big dogs like PwC’s might end up being cor­rect, but as of today, there are way too many ​“will”, ​“might”, ​“could” and ​“has the poten­tial to”. It will be hard for you to find a seri­ous study (not an adver­to­r­i­al) com­par­ing pro­duc­tiv­i­ty before and after. Until then, it’s all aspirational.

Buu­u­ut…

At the same time, we can’t ignore the fact that all of these large scale foun­da­tion mod­els are eff­ing amaz­ing. It’s much more inter­est­ing to research con­ver­sa­tion­al­ly via Per­plex­i­ty than to enter stuff on Google (more on that in a minute) and have to find your way around a dozen arti­cles. Deal­ing with spread­sheets, sort­ing data, com­bin­ing options, research­ing, test­ing, proof­read­ing. The appli­ca­tions are almost count­less. We use it every day for pret­ty much any­thing except, sur­pris­ing­ly(?), con­tent generation. 

Any­one with­out a tech­ni­cal back­ground can start join­ing dots with no-code plat­forms like N8N and build their own, very tai­lored appli­ca­tions that are unique for their needs and ways of work­ing. We’ll see AI being applied dif­fer­ent­ly to each ver­ti­cal. Sup­ply chain man­age­ment is already feed­ing huge amounts of data for AI to pro­vide improve­ments that can move the nee­dle through opti­miza­tion and per­for­mance and/​or cost reduc­tion. These are excit­ing and fun times, too.

So yes, to everything:

• This tech­nol­o­gy is here to stay
• We need to adapt to it and adopt it as soon as possible
• We need to learn how to nav­i­gate the non­sense tsuna­mi we live in
• And we shouldn’t expect it to be a life-chang­ing tech that will help you con­quer your market

It might be. We just don’t know yet. And the mon­ey you have to put to see if/​how it works for you might still be bet­ter invest­ed elsewhere.

Pre­dic­tions & con­tra­dic­tions 2: Search Engines are dead

Some even say con­tent strat­e­gy is dead. 

Check this screen­shot (source) from Ryan Law. Hub­spot, the absolute king of con­tent, the coin­ers of the ​“inbound” term, are wit­ness­ing their organ­ic traf­fic plum­met from a peak 6M ses­sions to 800k. 

There are many fac­tors that con­tribute to this down­fall of content.

1. Google sucks. Big time.
2. Also Google doesn’t care about you, your con­tent, if it’s good and rel­e­vant, and if it should, pure­ly based on its mer­its, be shown to the end vis­i­tor. It cares about push­ing ads, because they rep­re­sent +70% of its rev­enue. It’s not mak­ing more mon­ey because it’s bet­ter, or peo­ple use Google now more than ten years ago. It’s just because it’s more inva­sive and expen­sive, all at the expense of your content’s best interest.
3. Add AI research to the mix. Top of fun­nel con­tent tends to be very eas­i­ly and impar­tial­ly resolved by Chat­G­PT, Per­plex­i­ty, Claude, or what­ev­er your favorite chat­bot is. Users are los­ing a rea­son to drill down into the answers and vis­it a website.
4. And very much in tune with the ​“two oppos­ing ideas can be right” mantra, AI is also mess­ing with every­thing as com­pa­nies are irre­spon­si­bly falling into spam tech­niques of all sorts and pub­lish­ing lit­er­al­ly mil­lions of pieces of crap, auto­mat­ed con­tent on a dai­ly basis. The Dead Inter­net The­o­ry is no longer a theory. 

And yet! Noth­ing is ever that simple.

• Tim Soulo from Ahrefs shows us where the miss­ing traf­fic went. So Hub­spot is screwed by the cur­rent state of things, yes. But it’s also los­ing fights with oth­er brands doing a bet­ter job at con­tent, fair and square.
• Peter Capu­ta from Data­box shows us how traf­fic gen­er­at­ed from AI search results are mean­ing­less.
• And Mr. Law comes back to rethink his own post and bring some sense into the discussion.

And once you read it all, the con­clu­sion is simple: 

The days of hav­ing ​“easy-ish” #1 key­words are over. Google is not a reli­able chan­nel. We’ve known that for a long time. Any busi­ness that’s based on some­one else’s busi­ness is aware that’s a lia­bil­i­ty. Oth­er than that, noth­ing changes. Good con­tent is a must-have. It’s just busi­ness. And sales. And rela­tion­ships. And you need to have a voice, and com­mu­ni­cate it effec­tive­ly, and have some shit to say. The old black. The always black.

SEO will get hard­er but will still be there. Incen­tives are a very tricky thing, and you can rest assured that Google doesn’t want you to take your mon­ey else­where, so this will con­tin­ue to change. 

It’s not ​“the death of…” any­thing. It’s just ebbs and flows. 

Sim­ple. Qui­et. Good work.

I firm­ly believe that for SMBs the only way to grow is through con­sis­ten­cy. There’s only so much you can do at the same time. And tech­nol­o­gy can reach incred­i­ble veloc­i­ty, but our abil­i­ty to adopt and learn new tricks is lim­it­ed. No more pre­dic­tions. Just tips, hop­ing they’re use­ful for you this year:

• Take ​“dig­i­tal trans­for­ma­tion” (what­ev­er that means, which we’ll dig into next month) seri­ous­ly. Not as a start and end project, but as a prac­tice. It’s not about trans­for­ma­tion, but about prac­tice, adop­tion and evolution.
• Don’t let tech folks sell you bloat­ed tools you don’t need, or mag­i­cal solu­tions that promise things they can’t prove. Demand case stud­ies and take the time to make sure that’s exact­ly what you’re look­ing for.
• Pick one or two things. Try them on pilot pro­grams. Mea­sure results. Then scale. Don’t get into a 2‑year ​“any-ERP-brand”inte­gra­tion night­mare just because ​“they have new AI tools” or ​“my com­peti­tor has it”. Some­times your com­peti­tors are just morons and you’re bet­ter off doing nothing.
• And build great con­tent. Will it work on Google? I don’t know. But it will help you estab­lish your posi­tion in the mar­ket, cre­ate trust with your ide­al cus­tomers, and pro­vide tools to your sales team to get results.

It’s bud­get time!

Quick fig­ures in case they can help as you jot num­bers for the year:

• Your mar­ket­ing bud­get should stay around 10% of your rev­enue. If you’re bold­er or grow­ing fast, that’s great. If you had a dis­ap­point­ing 2024, that’s even a stronger rea­son to pro­tect that 10%.
• Not hav­ing a mar­ket­ing bud­get or hav­ing one under 7% is equal­ly use­less to gen­er­ate business.
• Rough­ly 40% of that bud­get goes into salaries (in-house, free­lancers, out­sourced, it’s the same thing — people’s time you have to pay for)
• Sales and tech­nol­o­gy expens­es are part of the mar­ket­ing bud­get. This leads to a big­ger con­ver­sa­tion, also for anoth­er day: Mar­ket­ing and Sales should be the same team.

Of course, if you need help mak­ing plans and set­ting pri­or­i­ties, you can con­sid­er our Blue­print ser­vice.

Life’s most­ly irony and con­tra­dic­tion. And Nor­man Fuck­ing Rock­well showed it masterfully. 

What would his paint­ings look like today?