Notes
Notes I write when a question comes up twice.
These are working notes, dated in 2026, written when the same problem showed up on more than one account. They are not a magazine. If a question has only come up once, it is still a conversation, not a page.
Start an enquiry What I do
12 Feb 2026
The enquiry form is usually the problem
I have a habit that annoys people who wanted a media conversation. Before I talk about budgets I open the site on my phone, on the train if I must, and I try to complete the enquiry. I use a number I can see go through. I use an email I check. I do not log in as admin. I do the thing a stranger does after seeing an ad at lunch. A surprising share of “need more traffic” conversations end about ninety seconds later, with a spinner, a validation error in English the visitor did not write, or a field labelled “Tell me about your project” that is required and has no hint of how much to write.
Forms fail in ordinary ways. Too many fields is the famous one. Asking for a company name, a job title, a budget range, a timeline, a phone, two emails, and a paragraph is a way of saying you only want enquiries from people who already trust you. Most of the businesses that hire me do not have that luxury yet. They have a useful service and a cautious stranger. The stranger will give a name, a way to reply, and a sentence. Everything else can wait for the call. I have watched clinics lose appointments to a “preferred time” picker that did not include evenings, and trades lose jobs to a file-upload that only accepted desktop PDFs.
Then there is the thank-you page that nobody wired. The form posts. The visitor sees a generic heading. Analytics records a page view if you are lucky, and an “enquiry” if someone once set a goal in a hurry. Meanwhile the inbox rule is broken, the message lands in spam, or it lands with a partner who is in meetings until Thursday. From the visitor’s side the business has gone quiet. From the business’s side there is no lead. Paid media, in that shape, is a way of buying silence at a higher volume.
Phone paths are part of the same problem. A number that is a picture on mobile, a click-to-call that dials a general line nobody picks up after 5pm, a voicemail box that is full. I treat a completed call as a conversion only when I have seen it connect to a human or to a recorded line that is actually checked. Until then, “calls” in a dashboard are a rumour. For F&B and clinics this matters more than the form. People still want to speak. If speaking is harder than tapping an app, they will tap the app, including the competitor’s.
What I do in the first fortnight, almost regardless of the brief, is make this path boring and measurable. One primary action on the money pages. A form that a tired person can finish. A notification that arrives where someone will see it within an hour on a working day. A thank-you page that is a goal, and a weekly glance at the actual emails or bookings so that the number and the reality cannot drift apart. Then, and only then, I am willing to talk about spending more to send people into that path.
Owners sometimes feel this is beneath the strategy conversation they paid for. I understand the feeling. The submit button is still where the money falls out. Raising spend on a broken enquiry path is a way to feel decisive. Fixing the path is slower to brag about and faster to cash.
9 Mar 2026
Branded search: when you're paying for a customer you already had
Branded search looks wonderful in a report. The click-through rate is high. The cost per click is friendly. The conversion rate makes every other campaign look careless. Someone, often a person who left the business two owners ago, set it running because it was the easiest way to make a dashboard blush. I have sat with specialist clinics and professional firms who were proud of that campaign, and I have had to be the person who explains that a large share of those conversions were people who already knew the name and would have booked through the organic listing sitting underneath the ad.
That does not mean branded search is always a waste. Competitors bid on names. Directories bid on names. Delivery platforms bid on restaurant names and then charge a commission for the privilege of returning a customer you already cooked for. If you disappear from the paid row, someone else will stand there with a similar colour and a hungrier call to action. Protection has a price. The question is whether you are paying protection money or buying a costume of performance.
The way I tell the difference is dull. I look at the organic ranking for the name. I look at what sits above it: maps, competitors, platforms. I look at the search volume, which for many Singapore SMEs is small enough to count in a morning. I look at the landing page. If the ad is the only reliable way to reach a page that actually takes a booking, the problem is the page, and the branded campaign is a crutch. If the organic listing already goes to a working page, and the ad is duplicating it, I cut the bid down to the level that keeps a presence without vacuuming the budget.
Then I have to talk about the drop. This is the part agencies dislike, because the first report after the cut looks like a failure. Conversions fall. Cost per conversion may rise on the remaining campaigns, because you have removed the easy ones. I write the paragraph before the call. I say: you are no longer paying to shake hands with people who were already walking in. The number that matters now is new patients, new tables, new mandates — the ones who did not know the name last month. If I cannot show a plan for those, I have no business touching the branded campaign at all.
There is a Singapore-specific wrinkle. Many buyers search in English mixed with place names: a neighbourhood, a mall, a MRT stop. Brand plus “Bugis” or brand plus a dish is not the same as the naked brand. I split those. A person adding a place or a problem is closer to a choice. A person typing only the name is often navigating. Treating them as one campaign is how the costume gets sewn.
If you take one thing from this note, take the habit of asking “would this person have found me anyway”. It is an unfashionable question because it makes a successful-looking campaign smaller. It is also how you stop renting your own reputation from a platform by the click. I would rather a quieter dashboard and a budget that can fund a page for the condition, the product, or the dish someone does not yet associate with your name.
21 Apr 2026
Reporting that fits on one page
I print the monthly report. Not always onto paper, though I have, on purpose, so that I can see whether it survives the printer. The test is whether a busy owner can read it on the MRT between Bugis and their stop and arrive at the call already knowing the argument. If the report requires a login, a filter, and a guided tour, it is not a report. It is a product demonstration. I left those behind when I opened this practice.
The page has a short head: spend, enquiries, cost per enquiry, and a close or booking note if the client shares one. It has two or three changes made that month, in sentences, not in a changelog. It has one recommendation. It has one thing I am deliberately not doing, so that “why isn’t this on TikTok” has an answer in writing before the call begins. If tracking is broken, that fact sits at the top, in ordinary language, because a pretty number under a broken tag is worse than a blank.
What stays off the page is the hobby material. Session graphs with no destination. Follower counts. Video view rates for a business that does not sell through video. Heat maps of the about page. Device breakdowns that have not changed in a year. I look at some of these in the tools, the way a mechanic looks at a bay that is not the job. I do not bill you to watch me look, and I do not staple the looking to the decision document.
People worry that a short report means I am hiding something. The opposite is true. A long report is an excellent place to hide. You can bury a wasted campaign under eight screens of “insights”. A page has nowhere to put a failure except in the sentence where it belongs. If branded search was a costume, the page says so. If the month was quiet because the offer is seasonal, the page says so. Quiet is allowed. Quiet explained is useful. Quiet decorated with charts is how retainers go to sleep standing up.
The call then has a job. I am not walking through the page. You have read it, or I wait five minutes while you do. The time on the call goes on the decision: move the money, change the page, pause, hire, wait. Thirty to forty-five minutes is enough if the document did its work. If a call regularly runs to ninety minutes, the report is doing too little or the decision-maker is not in the room. I will say that. I will not add slides to make the extra hour feel designed.
I have been asked to plug the page into a live dashboard so that leadership can “self-serve”. I have never seen that end well for a twelve-account practice or for an SME. Dashboards are refreshed in anxiety, usually on a Thursday, usually after a single bad day. The monthly page is a pace. It is also a promise that I will look at the account as a whole, not as a spike. If you want a live view of spend and enquiries I will show you where to find it. I will still write the page. The page is the work. The dashboard is a window.
18 May 2026
Why I cap the practice at twelve accounts
Twelve is not a brand flourish. It is the number at which I still know, without opening a tab, who answers the phone at each business, which landing page is the money page, and which specialist I would call if paid media needed an emergency pause on a Thursday. Past twelve I would be inventing a layer I left the industry to avoid: a junior to “watch the accounts”, a shared inbox, a weekly internal meeting about clients I no longer see. I opened Periodbenitoagency in 2026 so that the person on the call would be the person doing the work. The cap is how that sentence stays true.
There is arithmetic underneath the sentiment. A monthly hold is one report and one call, plus the actual channel work. A build is heavier. Diagnostics are short and dense. If I mix those shapes carelessly I run out of Tuesdays. Twelve active engagements, with at most a couple in a heavy build at once, is a week I can still do well from Bugis without pretending I have a studio. I have tried the other way, on the agency side, and I remember the week when I could not have named the conversion action on an account I was notionally leading. That memory is the policy.
The cap also protects the client when I am away. With twelve, I can give notice, pause what can wait, and brief two specialists on the accounts that spend enough to need a human during a fever or a family week. With twenty, that briefing becomes a fiction. Nobody actually holds twenty contexts in their head, and the people you hire to pretend otherwise will not have been in the sales call where the offer changed. Continuity is the product. Continuity does not scale by adding names to a Slack channel.
I am asked, often, what happens when a thirteenth interesting brief arrives. I say no, or I say later. I keep a short waiting list. I do not run a drip sequence to keep the thirteenth warm with content they did not ask for. If the work is urgent, I will try to point at another practitioner. If it can wait until a seat opens, I will say roughly when. Losing a fee is cheaper than taking a seat I will resent by month two, and cheaper than the client discovering that I have become an account manager of myself.
There is a vanity version of this story in which twelve is a luxury signal, a velvet rope. I dislike that reading. Some months I have fewer than twelve because I have finished a build and I will not fill the gap with a bad-fit retainer. Empty capacity is allowed. Filling it to make the practice look busy is how the cap dies. I would rather a quiet August and eleven accounts I can still describe than a full calendar of logos I cannot defend.
If you hire me, you are hiring a person who will still be on the account in six months, who will remember the decision from April, and who will not introduce you to “your new strategist” because the old one was promoted into air. That promise is inexpensive to print and expensive to keep. Twelve is the machinery of keeping it. When I raise the number, if I ever do, I will write another note, and I will owe you an explanation that is better than “demand was strong”.
30 Jun 2026
Long sales cycles and the patience budget
A B2B firm selling into plants, contractors, or regional buyers does not have a thirty-day sale. The committee meets. The specification sits. A champion goes on leave. Four months is ordinary. Six is not a scandal. And yet the marketing conversation is still held over a monthly dashboard, as if the calendar of the report were the calendar of the customer. I have watched competent search programmes declared failures in week six because the pipeline had not yet coughed. The programme was not the problem. The patience was unbudgeted.
I use a phrase internally, and now here: a patience budget. Before I spend on a long cycle, I write down how long a serious enquiry should take to become a conversation, and how long that conversation usually takes to become a fee. I take those numbers from sales, not from a blog. Then I set the first useful reading of the marketing work beyond the length of that cycle, or at least far enough in that a handful of enquiries could have matured. Until that date, I watch leading indicators: qualified enquiries, sales-accepted leads, replies to the sequence. I do not throw a party or a funeral over closed revenue in month one. Closed revenue in month one, on this kind of account, is mostly coincidence.
Leading indicators need discipline or they become vanity in a new hat. A “qualified enquiry” is one sales will pick up. If sales will not pick it up, it is not qualified, and no amount of scoring will make it so. I would rather eight enquiries a month that a salesperson returns to than eighty that they mute. Content for a long cycle has the same test: will a salesperson send the page. If they will not, the page is a brochure with a URL. I keep writing until they will, or I stop the programme and say so.
Email sequences exist because four months is a long time to leave a person with nothing but a memory of a PDF. They should be few, useful, and easy to reply to. A fortnightly essay that nobody asked for is not nurture. A short note that answers the question the last call actually stalled on is nurture. I write those from sales calls, which is why I ask to listen to them in the first fortnight. The sequence is not a marketing asset. It is a courtesy with a job.
The other half of patience is stopping. A long cycle is not a licence to run a bad account forever on the grounds that “B2B takes time”. If qualified enquiries are not appearing by the date on the plan, I change the page, the offer, or the channel, and I say that the hypothesis failed. Patience without a review date is inertia. I put the review date on the first plan so that nobody has to be the villain for calling time. The villain is the missing date.
If you sell slowly, hire someone who can stand in a quiet month without inventing a stunt. Stunts reset the clock. They also train the organisation to expect fireworks whenever a dashboard looks flat. I would rather a dull sequence, a short list of terms, and a sales team that trusts the form. That mix does not photograph well. It does fill a pipeline, which is the only picture that matters once the patience budget has been spent.
If a question of yours is not on this page, send it
If it comes up twice, it may earn a note. If it comes up once, it still deserves a reply within forty-eight hours on a working day.
Start an enquiry
