Small Business Marketing Budget: Splitting Spend by Channel

A workable small business marketing budget starts from the number of jobs you want, not from last year's revenue. Write the target, split the money into core spend, learning spend and slack, read each platform's budget unit the way that platform defines it, and review the split once a month against your own enquiry counts. If you run a print shop that serves walk-in customers and local businesses ordering flyers, banners and business cards, that routine fits your situation. A ready-made percentage split is not part of it, because the right split depends on your target and your own conversion records.

Start with the order you need, not the revenue you had

A budget built as "a slice of last year's takings" shrinks in a slow season, exactly when more enquiries are needed. It also says nothing about which channel deserves the money. A budget built backward from a target asks two questions instead: what extra order volume do I need this quarter, and what am I willing to pay to win one?

Working backward from jobs

Take one target and one period. A print shop might say: "Over the next ninety days I want more repeat orders from local businesses, and I can commit a fixed amount per month to paid channels." That sentence names the outcome and the ceiling. It does not yet say how an enquiry is counted, which is the third thing to settle before any money moves.

Defining what counts as an enquiry

A phone call, a quote request, a message, a walk-in with a file on a USB stick: each is a different event. Decide which ones count as one enquiry and write the rule down. Without it, two channels can report the same customer twice, and the monthly numbers will not add up.

Putting a value on an enquiry

Google's Ads help page on conversion values says values help you measure and optimise the true business impact of campaigns, and let you see the total business value Google Ads generated rather than only the number of conversions. It describes two ways to assign them. Different values per conversion are more representative of the real value of each transaction but can take more work to set up. The same value for every conversion is easier to set up but may be less representative. The page adds that without dynamic value code in the tracking tag, the default value you enter is applied to all conversions.

Practical suggestion: a print shop sells jobs of very different sizes. Start with one default value for a typical quote request, compare it with real invoices after a month, and refine it later.

Core, learning and slack: a three-way split

Divide the monthly budget into three parts, each with a different job. Core spend keeps working whatever you decide this month: profile upkeep, regular content, asking for reviews. Learning spend funds one unproven channel or message at a time. Slack is held back for a seasonal rush, a fast fix or a channel that surprises you.

Table: three parts of the budget and how each is judged. The review rhythm is a practical suggestion.

How big each part should be

No source gives a ratio, and a made-up one would mislead. Practical suggestion: a shop whose orders arrive through repeat customers needs a larger learning part to find new ones. A shop with one big seasonal peak, such as back-to-school printing or an annual event poster run, needs a larger slack part. Whatever the sizes, having all three parts matters more than their proportions in the first quarter.

Moving money between parts

Money can move, but only with a written reason. Learning money moving into core means a test earned its place. Slack money moving into learning means a core channel weakened or an opportunity appeared. A transfer without a reason leaves the monthly meeting unable to answer what the money bought.

What "budget" means on each platform

The word covers different rules on different platforms. Read the platform's own definition before putting a number in the table.

Google Ads: the average daily budget

Google's help page defines the average daily budget as the average amount you set for each ad campaign on a per-day basis. To turn a monthly figure into a daily one, divide by 30.4, which the page describes as the average number of days in a month, 365 divided by 12. Its example turns 304 dollars a month into 10 dollars a day. On a particular day, spending can reach two times the average daily budget, and the page gives that as its general rule. In a month, spending is limited to 30.4 times the average daily budget under the same rule. The page also says on some days you might not reach the average and on others you might exceed it. You can edit the budget at any time, and shared budgets let one budget cover several campaigns.

Meta: daily and lifetime budgets

Meta's Marketing API documentation defines a daily budget as the average amount you are willing to spend on an ad set or campaign each day, and a lifetime budget as the amount you are willing to spend over the entire run. The page says a daily budget paces spending across each day, while a lifetime budget paces spending over the ad set's lifetime. On days with better opportunities, up to 25 percent more than the daily budget may be spent. With a lifetime budget you are not charged more than that budget for the ad set's results unless you change delivery settings. The documentation sets budgets at the ad set level and sends campaign-level budgets to a separate guide.

Platform limits can change, so check the current pages before relying on a figure.

Why the difference matters for the table

If you write "100 a day" in the worksheet for both platforms, you have written two different promises. In Google Ads it is an average with a two times daily ceiling. In Meta it is an average that can run up to 25 percent over on good days. In the worksheet, record the platform, the budget type and the monthly ceiling you compute.

Paid and organic run on different clocks

Paid channels are read after the campaign runs. Organic work has a longer and less certain horizon. Google's SEO starter guide says some changes might take effect in a few hours while others could take several months, and that you likely want to wait a few weeks to assess whether your work had beneficial effects. It gives no further number.

Where this leaves organic work

Organic work belongs in core spend, because a quarterly check matches its horizon. Judging it on a one-week window with the same impatience as a paid test will cut good work early.

Word of mouth and promotion

The same guide says effectively promoting new content leads to faster discovery, and it describes word of mouth as an effective and lasting way to promote a website. It also warns that you can overdo promotion and harm your site. For a print shop that depends on repeat clients, this is an argument to keep a steady review and recommendation routine inside core spend.

Paid work and a fair test

A paid channel earns its place through a written success measure and a date. Cut it before that date only for an operational reason, such as a broken tracking tag, never because a few days looked flat.

The smallest meaningful test per channel

Every channel needs a smallest test that is big enough to learn from. Because no source names a universal amount, the practical way is to define each channel's unit, signal and horizon in a table and let your own numbers decide.

A test needs an end date and a measure

Write both before spending. The success measure should be a count or a value you can read from your own records, such as the number of quote requests. An end date prevents a test from drifting into a permanent expense.

What to do when the test is too small

If the planned amount cannot produce enough events to read, either extend the period or narrow the question. Do not enlarge the budget just to feel safer. A narrower question, such as one product line, is cheaper to answer than a broad one.

The monthly money meeting

Thirty minutes once a month is enough for a small shop. The owner and, if there is one, the person who runs ads sit down with the worksheet. Reviewing daily is too noisy and reviewing yearly is too slow.

Agenda

1. Last month's target against actual enquiries.
2. Each channel's spend and enquiries.
3. The status of each learning test: running, finished or stopped.
4. Whether slack was used, and why.
5. Next month's decisions: grow, cut or hold.

Decision rule for grow, cut or hold

If a channel met its written success measure and its end date has passed, it moves into core and keeps its budget. If it missed the measure and there is no new reason to extend, it is cut. If it met the measure but money is limited, it holds at the same amount. A channel without a written measure and end date can be neither grown nor cut, because the meeting has nothing to judge it by.

The record the meeting produces

Leave with the updated worksheet, one decision sentence per channel and the date of the next meeting. Without those three items the meeting has not changed anything.

Allocation worksheet

Table: blank worksheet. The dots are for your own amounts.

Filling it in

Record actual spend from your own books, not only from the platform dashboards. Count each customer once even if two channels claim them. Compute the monthly ceiling for Google Ads as 30.4 times the average daily budget, and for Meta take the lifetime figure or multiply the daily budget by the number of days the ad set runs.

Example scenario: a print shop with walk-in and business clients

Example scenario: A small print shop sells copies and posters to walk-in customers and flyers, banners and business cards to local businesses. It has relied on word of mouth and a business profile. It wants more orders from local businesses within three months.

Setting the target and the parts

The owner writes the target as more repeat and new orders from local businesses over ninety days. Core spend covers profile upkeep, a monthly sample photo post and asking every business client for a review after delivery. Learning spend covers one search ad campaign aimed at business printing. Slack is kept for an annual event poster rush.

What the first review shows

At the end of month one the profile shows more directions and calls. The search campaign has run but the quote-request count in the shop's records is lower than the platform's conversion count, because some quote requests came by phone and were never entered. Decision: the campaign holds, and the owner starts logging every phone enquiry with a source question.

What the second review shows

In month two the log is complete and the search campaign meets its written measure at the end date. The owner moves it into core and opens a new learning test for banners. Slack is untouched.

Third review and the quarter's close

At the end of month three the owner compares the quarter against the written target. The banner test, run with its own success measure and end date, did not meet the measure, so it is cut with the reason recorded. Slack is carried into the next quarter because the event poster rush has not arrived yet. The worksheet now has three months of rows, which makes the next quarter's split an edit rather than a rebuild.

Lessons

• Platform counts and shop records need to be compared every month.
• A test that cannot be measured is not yet a test.
• Moving a channel between parts is written down.

When the budget is very small

With a small budget, testing every channel at once leaves each one underfunded. Starting with a single paid channel is safer, and the choice can be made with three questions.

1. Does your service answer something people actively search for?
2. Can you set up counting for this channel before spending?
3. Does the channel's reading time fit inside your budget period?

If all three answers are yes, the channel is the first test candidate. If one answer is no, fix that problem first. Profile upkeep and website content cost time rather than ad spend, so they stay in core regardless of the three questions.

Why one channel at a time

One channel makes cause and effect easier to read. If two channels start together and enquiries rise, you cannot say which one worked. The second test begins after the first test's end date, so the learning part is never split across two unreadable experiments.

When to add the second channel

Add it once the first channel has been judged and a decision recorded. If the first moved into core, the learning part is free for a new test. If it was cut, the remaining money can move to the second test, with the transfer reason written down.

Reading platform numbers against shop records

A dashboard shows what the platform could see. A print shop also receives phone calls, walk-ins with files and messages that never touch a tag. Comparing the two sources each month keeps the worksheet honest.

Log every enquiry once

Keep a simple log with date, type of enquiry, source answer and outcome. Ask each new customer how they found you, and give the staff a fixed list of answers, since free text is hard to total. The log is the source for the worksheet's enquiry column.

The gap between the two sources

Google's conversion value page says the conversion value/cost column lets you track return and compare keywords, ad groups and campaigns, so you can adjust bids, budgets and targeting by hand. That data covers conversions the platform recorded. Enquiries that reach you by phone without a recorded conversion appear only in your log. When the platform count is lower than the log, the gap is not an error to hide but a reading to explain at the meeting.

Counting a customer once

If a customer saw an ad, checked the profile and then phoned, three sources may claim them. In the worksheet, count the person once and note the channels that touched the journey. Attribution models are a larger topic, and the worksheet only needs a rule that prevents double counting.

Checklist before you split the budget

• Is the target one sentence with one period?
• Is it clear what counts as an enquiry?
• Is a default conversion value chosen?
• Are core, learning and slack separate?
• Does each platform's budget type appear in the worksheet?
• Does each test have a success measure and an end date?
• Is organic work given a longer horizon than paid work?
• Is the first monthly meeting in the calendar?

Misreadings to avoid

Copying a percentage from the web

A table that says "so much to search, so much to social" has no link to your target or your conversion data. It is a guess presented as a rule.

Treating a day of spend as a verdict

Google Ads can spend up to two times the average daily budget on a single day under its general rule, and Meta can spend up to 25 percent over. One day's spend says little about a channel.

Judging organic work on a paid timeline

The SEO starter guide's own range runs from hours to months, so a one-week verdict on organic work is premature.

Leaving slack untouched for too long

Slack that sits unused for a long stretch may be better placed elsewhere. Slack that is spent every month is no longer slack. Use it on a recorded reason.

Skipping the value

Without a conversion value, Google Ads shows counts. Counts are useful but treat a small poster order and a large banner order as equals.

Stopping the budget from drifting back to revenue

After a good quarter it is tempting to set next quarter's budget as a share of the takings again. Resist that. Start the next worksheet from the new target, carry over the channels that moved into core, and treat the previous quarter's totals as a record, not as the starting point.

Related budget decisions

Channel-level budgets sit inside the split. How much to put behind organic search is covered in the SEO budget for small business guide. Link and outreach spending has its own plans in the backlink budget guide and the sponsored post budget plan. For choosing channels by audience and measurement, see the website promotion channel table.

Frequently Asked Questions

How should a small business split its marketing budget?

There is no ready percentage. Write the target, split the money into core, learning and slack parts, and review the split monthly against your own enquiry counts.

Does Google Ads spend my daily budget in full every day?

No. Google's help page says that on some days you might not reach your average daily budget and on others you might exceed it. Its general rule allows a single day to reach two times the average.

What is the difference between a daily and a lifetime budget on Meta?

Meta's documentation says a daily budget paces spending across each day, while a lifetime budget paces spending over the ad set's lifetime. On better days up to 25 percent more than the daily budget may be spent.

How long should I wait before judging SEO work?

Google's SEO starter guide says some changes might take effect in a few hours and others could take several months. It suggests waiting a few weeks to assess whether your work had beneficial effects.

Do I have to set a conversion value?

No. Google's help page says values help measure real business impact, and it offers a simpler same-value option. Without a value you see only counts of conversions.

How much should I keep as slack?

The documentation has no statement on the amount. Having a slack part and recording the reason whenever you use it matters more than its size.

Conclusion

A channel budget is a decision routine rather than a percentage. State the target, define what counts as an enquiry, split the money into core, learning and slack, read each platform's budget type as that platform defines it, and review once a month with a written decision for every channel. Treat organic work on its own clock and keep the record of why money moved. Platform pages are the source for the budget rules used here: Google Ads average daily budgets (Opens in a new tab), Meta budgets (Opens in a new tab), Google Ads conversion values (Opens in a new tab) and Google's SEO starter guide (Opens in a new tab).

If you want help producing the articles and pages that feed the organic part of the split, see the content writing service.