[ttul]

People inexperienced in online adversities often think that they can pay for conversions. That is not the case. You’re paying to get your message out. Very very few people will immediately act because of an ad. But if you hit them enough, they may later convert.

Tracking those late conversions is imprecise. This does not mean advertising is useless. But it is not generally a game for those with a tiny budget or a short time horizon.

[ggg9990]

Disagree. I had a startup selling a piece of software that immediately solved a major problem for a small niche of users. Paid for conversions on common search terms and they were ROI positive since each click converted at 25% to signup, and then the product spread around the companies where one user adopted. Sold the company for >$10m.

[282883392]

Sounds like that was more of great product than a successful advertising campaign, even if it did help.

[ggg9990]

The product always needs to be good. But targeted ads such as on Google allow entire categories of businesses to exist. My company would have never been started without Google AdWords - there would have simply been no cost effective way to find the users who needed the product.

[rsync]

"The product always needs to be good. But targeted ads such as on Google allow entire categories of businesses to exist."

With regard to niches, my experience has been the opposite.

If your niche is small enough, you can't properly bid on adwords for rare keywords because they are flagged as "low quality" and you are forced to either abandon those keywords or pay 10x for them to show.

This, of course, defies the entire raison d'etre of adwords - the whole purpose of which is (presumably) to bid on a keyword that targets that one rare person that searched for that.

But it can't be done - if only 5 or 10 people search for a phrase or keyword every day, it cannot be bid on properly.

[ggg9990]

I don’t know about the 5 to 10 search a day volume level - I was quite a bit above this. But I agree that if they don’t serve that volume level well, it’s sad and aggravating. That should be something that can really help people who are out of other options. Might it be a privacy issue? If an ad can be targeted to that level it can be used to address a single individual.

[sah2ed]

> If an ad can be targeted to that level it can be used to address a single individual.

Alec Brownstein did exactly that by ad targeting his favorite creative directors' names some years back in order to land a job. It worked.

https://mashable.com/2010/05/13/job-google-ad-words/#GvW9atl...

[jypepin]

advertising is about telling people about your product. If you don't have a great product, it won't sell, no matter how much ads you put out there.

[ggg9990]

This is usually true but sometimes not. For example products like Grey Goose vodka make billions of dollars despite being indistinguishable from other quality vodka even by self-professed experts.

[mseebach]

It's probably splitting hair at this point, but Grey Goose isn't bad vodka, it's just no better than other "good" vodkas (whatever "good" means in context of a product whose defining characteristic is the absence of characteristics). If GG was a foul-tasting vodka, no amount of advertising would save it.

[jypepin]

this ^ I didn't say you need a better product, or the best product, but you need a great product. If people hear about your product, and they think it's great, then they'll buy it. GG vodka is a great vodka (not a vodka pro but definitely in the high end of consumer vodka I believe), and coupled with their advertising they sell a lot of vodka.

[nathan_f77]

Wow, that's awesome! I have a product that also solves a problem for a small niche of users [1], but I think my niche is a bit smaller than yours. I recently increased the AdWords budget and I'm trying a lot of different ads to see what works. Anyway, thanks for sharing your story, that's really encouraging.

[1] https://formapi.io

[dylanha]

How did you find the niche? Any reading material I should see?

[ggg9990]

I had a CS degree but was working outside the field in pharmaceuticals. I was having the problem myself and wrote a crazy Excel spreadsheet with tens of thousands of lines of poorly written copy pasted code to do it. I knew that there would be a market for it so I contacted an old classmate of mine who rewrote it much better as a web service.

That is why I think CS and software engineering should be part of every college degree, so that people can see the opportunity for software where it exists. I wasn’t good enough to be a programmer but I had enough knowledge to see an opportunity whereas most people thought “hm, this is just how this is done.”

[nathan_f77]

Reminds me of this AMA answer by @patio11 [1]

> Here's an exercise you can do: do you understand what a life insurance agent does all day every day? Make it your mission for a week to do so, well enough to explain it to a close friend who has no access to your sources. All you have to do to learn this is read and make conversations happen. (People are happy to talk to you!)

I really need to take that seriously, and talk to people in a lot of different industries.

[1] https://www.indiehackers.com/forum/im-patio11-patrick-mckenz...

[nodamage]

Just curious how much the software cost?

[ggg9990]

Individual contracts around $50,000 per year per customer.

[nathan_f77]

And you were getting these customers from AdWords campaigns, instead of high-touch sales? That's incredible.

[sfifs]

There actually are ways like market mix models and multi touch attribution models to account for impact of those "late conversions" across multiple channels but requires fairly sophisticated data integration and modeling and is generally worth it only if you are optimizing millions of dollars in ad spend. The tools Google/Facebook currently provide are not really sophisticated enough for such work.

Source: I work on these techniques

[istorical]

I work in analytics in the sales and marketing department at a late stage startup. I was recently tasked with moving from just reporting to working in modeling and prediction.

A first step I thought of was to build some multiple variable models / apply multiple linear regression to try to measure the relative influence of our marketing campaigns and touches to a sale.

With your experience in this area - are there are particular models you recommend trying to build as a first step for someone who's more of a beginner in modeling and predictive analytics? Or business questions you find go the furthest within this specific domain of sales and marketing?

[KodiakLabs]

Could you possibly point to by resources on these models?

[sfifs]

While details tend to be proprietary, MMMs are well established and lots of academic papers online. For instance

https://web.archive.org/web/20120813035903/http://faculty.fu...

High quality multi touch attribution models are still fairly new and I've not seen much public literature on the designs. We've had to spend a fair amount of time in development ourselves vs. relying on existing methods.

[libertine]

Precisely - "marketers" these days don't grasp the concept of brand building. Just because you can measure directly results it doesn't mean the goal is to get a direct conversion - specially from a platform like youtube where people are actively consuming content - not buying stuff.

The idea of compound investment into brand building doesn't seem to exist.

I don't know if these guys thought that decades of TV advertising campaigns were made by clueless people, with no education, with no market research, benchmarks, that dumped millions to build reach and frequency to try to fill a small spot in the mind of consumers.

[Silhouette]

Precisely - "marketers" these days don't grasp the concept of brand building.

I'm not sure that's true. Some people are just working on a different scale.

Brand-building is fine if you're Coke or BMW or McDonald's. You have millions to spend, you can saturate channels, and then people might actually think of you in preference to your competition when they're in the market for a drink or a car or a burger.

On the other hand, if you're a bootstrapped startup and worrying about affording this week's ramen, ttul's "tiny budget and short time horizon" is probably all you've got. To a first approximation, the only thing that matters for online ads in this environment is how many people you can get to take some immediate useful action, even if it's just volunteering an email address that you can use to follow up, or sharing your site with someone else they know who might be interested.

Of course, in the longer term, brand awareness is valuable as well. But branding isn't worth squat if your business won't be around long enough for someone to remember it later.

[libertine]

I agree that the work of marketers differs from scale, but it doesn't justify bad decisions.

I don't agree that brand-building is fine just for major brands, where they fight for small % of a saturated, well defined, market share.

Like you said, it's all about scale, and actions of a bootstrapped startup should adjust to it - not to be spending thousands on Youtube campaigns (picking up this post example). If you want some immediate useful action you should focus on media and placements where you have a higher percentage of people that a near to take immediate useful actions - maybe a video platform to consume video content is not the best choice to start with.

Brand building can be as simple as the way you address your customers, how your company looks, how the buying/service experience is, etc - things that change the perception of something random and generic, yet foundational to any brand.

McDonald's was the place to get tasty food without having to wait sitting down for it.

[icebraining]

you should focus on media and placements where you have a higher percentage of people that a near to take immediate useful actions - maybe a video platform to consume video content is not the best choice to start with.

Where would that be for a candy subscription service? I mean, people looking to buy candy probably just want one packet right now, not ten spread somewhere over the next few months.

[wolco]

In a startup brand awareness p is extremely important perhap more so than coke. They spend millions of dollars because they are trying to reach billions.

The most useful thing I noticed some startups do is get involved in their niche community and spent money building brand awareness for a targeted group.

[danmaz74]

Branding isn't just about spending a lot on ads. I found lots of actionable suggestions even for a small startups in a book called "Killer brands" by Frank Lane.

[ttul]

This week’s ramen won’t come unless the ads you ran six months ago generated some form of brand recognition...

[Silhouette]

That's going to be tough if our hypothetical business has only been running for three months, though, isn't it? Everyone has to start somewhere, and all I'm saying is that in those early days, the need for immediate results will naturally dominate longer-term brand-building -- unless you're already heavily funded and don't have the same priorities, of course.

[DisruptiveDave]

I'd venture to say it's not the marketers acting on their own in many cases. Shit runs down hill. VCs want fast returns, CEOs want their investors to be happy about their CAC, marketers want to prove their value in this system. And EVERYBODY is obsessed with analytics, often as a poorly constructed crutch. You try pitching brand building in that environment.

[libertine]

But I understand that, so maybe it's a matter of choice.

What kind of brand you want, and set it right away from the beginning because it will define your media channel/production investment.

I think that brands who had strong branding advertising are still well positioned in people's minds, and it's indeed an investment that should be made with a massive ROI.

Now, it's a long term game. Doesn't suit the status quo like you well said.

[e_b]

The long term brand plan must be set by the CEO and in the example provided by the parent, the CEO only cares about keeping investors happy in the short term, alas no spend without immediately identifiable ROI.

[e_b]

Exactly this.

[gk1]

What marketers are you talking about? Who are "these guys"?

I work with marketers and I am one myself, and the concept of brand-building is as familiar as loops are to software developers.

No doubt you've encountered misguided individuals, but please don't generalize from specifics.

[libertine]

Anyone with a Google Adwords Certificate is a marketer these days.

> No doubt you've encountered misguided individuals, but please don't generalize from specifics.

When the community gives visibility to bad examples, specifics gain other dimension. So instead of blaming me for generalization, maybe understand the context.

I'm a marketer as well.

[dahdum]

Brand building investment exists today, but like all other channels unless you can prove the return you're just gambling. It's certainly not the best place to start if performance marketing is a viable channel.

[libertine]

I know such investments still exist - specially in well established brands where a more "classical" marketing approach is still used.

But small companies don't even know how to measure such returns, or can't afford to measure their return beyond sales and analytics data. That's why it will always be a gamble for them, specially like you said, it's not the best place to start.

[vidarh]

It's not just that they can't afford to measure their return, but that often you can't afford to spend money on marketing that doesn't have a direct return.

When I started consulting, I spent some money on marketing, but my big constraint was that if a given channel didn't at least break even on direct conversions, there was only so much money and time I could afford to spend on that channel for conversions down the road by increasing my brand: It's not just total lifetime returns from the advertising that matters, but how the timing of those returns affect cash flow.

If the time it takes to break even is too long, you might be bankrupt first, even if the potential lifetime return is fantastic.

Sometimes it costs a lot to not have the time and budgets to take a longer term view.

[tinymollusk]

I have no question that people who have built brands believe that it was successful. We can even look at "big brands" like Coke -- they sell sugar water + a brand -- surely that must be real.

My nuanced take is that a brand "exists" in the sense that we are able to understand it, what it represents, etc. What we do not know is if that's the best way to spend resources. All the money that went into Intel Inside -- what if they spent it on something else? No marketer can know that; the question as a marketer: "is this the best thing I can do with my resources?" Nothing about "building a brand" provides objective feedback or testable predictions about the usage of that dollar.

Suggesting that TV advertising can't be a scam because so many people have put so much effort into it is another fallacy. Old school TV advertising was done by people whose jobs dependent on the revenue stream -- even if branding were completely ineffective; if they got paid, I wouldn't consider that irrational.

Brands happen. They can have value, but we can only try to prove that value after the fact.

Direct response, on the other hand, tends to be less full of b.s.

https://conversionxl.com/blog/cro-vs-branding/

[petercooper]

This lesson struck me hard when I was working with a major conference organizer. Even upon surveying them, they'd struggle to establish how attendees had heard of the event first. For the majority of attendees there was no direct way to correlate ticket purchases with ad spend because almost no-one immediately buys a $1000+ ticket after clicking an ad :-)

[jackgolding]

Market research is the easiest way to solve these problems!

[ChristianGeek]

What if you provide a discount code in the ad that’s unique to the ad? Subsequent ads have slight lower discounts so there’s an incentive to use the code from the first ad they saw.

[petercooper]

Discount codes was one method used, the only problem was that at the $1000+ ticket level, it's often another department buying the tickets and they frequently don't bother to use the codes :-D

[icebraining]

Might work for some people, but the probability that I would even remember the first ad had a code - let alone the exact string - is minuscule.

[lozenge]

Not being able to find a discount you saw can be pretty frustrating. Less important in this case as employers often pay for these conferences.

[kraig911]

Uhh look up funnel marketing. There's a whole system in place that exercises and excels in exactly what you say we in advertising are inexperienced? I mean you can technically pay for views it's just not anything over 5% of impressions though... retargeting though works.

[jehlakj]

Indeed. People are subjected to feel more secure about a brand that is more familiar to them (e.g., Tide).

[ekianjo]

Indeed you mainly just buy awareness of your service. The road from awareness to trial can be quite long.