How to get growth moving again when it stalls

Your first signups, orders or enquiries came in, and then growth stalled: the weekly count went flat and nothing you can see changed. When you have only a handful of customers, a stall almost always has one of three causes: something broke, the one place your first customers came from ran dry, or you stopped doing the work that brought them. Here is how to get growth moving again: find which cause it is in about 2 hours today, then restart at a volume that shows you within 4 weeks whether it is working.

Do these three things

About 2 hours today, then 3 to 4 hours a day. Judge the restart after 2,800 messages, about 4 weeks, not after a quiet day.

Show the three actions in fullHide the full list

1. 1Today: count your signups, orders or enquiries for each of the last 8 weeks from your own records, not from visits. Then sign up, buy or enquire as a stranger, on your phone, to check nothing is broken.
2. 2Every day: send 100 personal messages to new people like your first customers, and give 20 answers in the kind of place your first customers came from. AI drafts each message; you add one true line about the person.
3. 3Every Sunday: write down what you did this week and what came in, by source, next to your best 4 weeks. Give next week's hours to the source that is bringing customers back.

Get help doing these

ChatGPT or Claude reads this guide and works through it with you.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

1. Count your signups, orders or enquiries for the last 8 weeks to see if growth really stalled

Write down how many signups, orders or enquiries you got in each of the last 8 weeks, counted from your own records. Growth has stalled if the first 4 weeks brought at least 6 and the last 4 weeks brought half as many or fewer. If the first 4 weeks brought fewer than 6, growth never really started, and you need a different fix.

Count the one result that makes you money, the same one you set up in conversion tracking. Take it from where the money or the person really shows up, and take out yourself, your friends and your family. Friends buy to be kind, and their orders hide the real trend.

If you run a SaaS: count new signups from your own database, and new paying accounts from Stripe's Billing overview. Its "New subscribers" number counts only paid subscribers who became active for the first time, with no trials. Also open "Subscriber churn rate", because you need it in step 3.

If you run a D2C store: in Shopify, open Analytics › Reports › "New customers over time" for first orders, and count all orders by week. Do not judge by visits. Between 21 and 23 September 2026, Shopify started taking bot visits out of its session reports by default, so your visits may have dropped even though no real shopper left. If your ads say sales fell, check that against Shopify's orders. Since 12 January 2026, Meta no longer reports its 7-day and 28-day "view" windows, so ad-reported purchases dropped for many stores with no change in real orders.

If you run a B2B company or an agency: count enquiries, replies that are not "no" and calls booked from your inbox, your sheet or your CRM. New clients come slowly at your size, so also count calls and quotes, which move sooner.

Why the rule uses 4 weeks, not one. With small numbers, an empty week often means nothing. If you get about 2 a week, you will get a week with 0 roughly 1 week in 7, by chance alone. Two blocks of 4 weeks smooth that out. If your first 4 weeks had fewer than 6, there was no growth to stall: follow why no one is buying your product instead, because its first fix is reaching 1,000 people who fit.

• Tool: Google Sheets, plus Stripe, Shopify, or your inbox.
• Time: 30 minutes.
• You will have: one row for each of the last 8 weeks, and the word "stalled" or "not stalled" written next to them.
• It worked if: every number matches your billing, your orders or your inbox, and the rule gives a clear answer.
• Common mistake: reading visits, likes or the ad dashboard instead of real signups and orders, or panicking over one empty week.

2. Sign up, buy or enquire as a stranger to check nothing is broken

Go through the exact path a new customer takes, on your phone and on a laptop, in a private browser window, all the way to the end. A sale or signup that fails quietly looks exactly like growth that stalled, and it is the fastest cause to fix.

Use a new email address you have never used on your site, and pay with a real card or a 100% discount code you delete afterwards. Do every step a stranger does. Then write down every change you made in the last 8 weeks, with its date: a new theme, an app, a release, a price, a form, a domain or an email tool. If the stall started the same week as a change, that change is your first suspect.

On your phone, then on a laptop, in a private window, with a new email address:
Open your site from your social bio link and from a Google search for your name. Does it load in under 5 seconds?
Sign up, check out or send an enquiry, to the very end. Did it finish without an error?
Did the welcome, verification or order email arrive within 5 minutes, in the inbox and not in spam?
Did the link in that email work?
Did the discount code or free-trial offer you promote actually apply?
Did you, the owner, get the notification for the new signup, order or enquiry?
In Google Search Console, do "Manual actions" and "Security issues" both say no issues?
Every change in the last 8 weeks, with its date: [theme / app / release / price / form / domain / email tool]

Real breaks are often small. In February 2026, a store owner on the Shopify Community saw her conversion rate fall from about 1.8% to 0.3–0.4%. On 9 March she found the cause: "The discount code was not applying in the cart drawer slider. Fixed it and it was like night and day."

Emails are the other quiet break. Since February 2024, Gmail has required every sender to have SPF or DKIM, two settings that prove an email really comes from your domain, and to keep spam reports under 0.3%. Since November 2025, Gmail has been stricter about mail that breaks these rules. If your verification or order email goes to spam, people sign up and never come back, so ask your email host to confirm both settings are on. Your email deliverability affects every email you send, not only marketing.

If Google search brought most of your visits, open the Performance report in Search Console, click the Date filter, choose the Compare tab and compare the last 3 months with the period before. In the Pages table, choose "Clicks Difference" to see which pages lost the most. Google's 2026 ranking updates ran on 27 March to 8 April and 21 May to 1 June (core updates), and 18 to 21 August and 24 September to 7 October (spam updates). If your clicks fell in one of those windows, read step 3's search row.

If you run a D2C store: add one test with your abandoned cart path: put something in the cart, leave, and check the reminder arrives. Also check the page loads fast on a phone. A slow theme or a heavy app can slow every page.

If you run ads, open the ad account's notifications for a paused ad, a rejected ad or a failed card before you change anything else. If people click and still do not buy, fix the ads that get clicks but no sales.

• Tool: your phone, a laptop, a private window, and Google Search Console (free).
• Time: 45 minutes.
• You will have: the test list with a pass or fail on every line, and a dated list of changes.
• It worked if: you finished the test on both devices, every email arrived in the inbox within 5 minutes, and anything that failed is fixed today.
• Common mistake: testing on your laptop while logged in as yourself. Most buyers arrive on a phone, as strangers, with nothing saved.

3. Write down where each customer came from and what you did each week

Make one sheet with two parts: every customer you have, with where they came from, and every week, with how much marketing work you did. Put them side by side. At this size, the cause is usually plain once you see both: one source brought most of your customers and then went quiet, or your own work dropped in the weeks the count went flat.

Customers and sources. For each customer, write one source. Take it from the answer to "How did you hear about us?", which is self-reported attribution, from the UTM tags on their first visit, or from the message thread where you met them. If you do not know, write "unknown" and ask them in step 4.

Weeks and work. For each of the same 8 weeks, count what you did from your sent folder, your post history and your community profiles: personal messages sent, follow-ups, answers given, posts, any launch, and any ad money. Be honest. Many founders go back to building once the first customers arrive, and the count falls a few weeks later.

Then find your cause in this table, top row first. Use the first row that fits.

Most founders with a handful of customers land on row 2 or row 3. A first source running dry is normal. Your friends and contacts run out, a launch day ends, and a post or comment that brought people stops being seen. It does not mean your market is gone.

If you run a SaaS: check how many of the recent signups did the main thing your product is for. If fewer than 4 in 10 did, people still arrive but stop before the fix, so raise your activation rate as well as step 5.

• Tool: Google Sheets, your sent folder, your social and community profiles.
• Time: 45 minutes.
• You will have: one sheet with every customer's source and every week's work, and one named cause.
• It worked if: every customer has one source, every week has a count of your work, and you can finish the sentence "Growth stalled because ___".
• Common mistake: blaming the market, the season or an algorithm before you count your own work. Check Google Trends for a seasonal dip only after your own numbers are written down.

4. Ask 10 recent customers and 10 people who stopped what brought them and what stopped them

Send 20 short personal messages today: 10 to your most recent customers, asking what brought them and what made them start, and 10 to people who signed up, added to cart or enquired and then stopped, asking what stopped them. Their answers tell you where to find more people like your first customers, and what to fix before you send more people.

Write each message yourself or let AI draft it from the template, then add one true line about what that person did. Ask one question and make it easy to answer. Do not sell. Write every answer into your sheet in the person's own words. When 3 or more people say the same thing, act on it: a place to find more people goes into step 5, and a reason people stopped gets fixed this week.

Hi [name], thank you again for [signing up / your order / working with us].Could I ask one quick thing? Where did you first hear about [product], and what made you start that day? One line is plenty. I'm trying to find more people like you.[Your name]

Hi [name], you [signed up / added [product] to your cart / asked for a quote] on [day] and didn't go ahead. That's completely fine.Could I ask what stopped you? Even one word helps: price, not sure what it does, didn't trust it, not the right time, or something else.I won't try to sell you anything. Thank you.
[Your name]

Keep it going: from today, send the second message to everyone new who stops, within 24 hours. If someone will talk for longer, book a 20-minute customer interview. If past customers have gone quiet, the win-back email asks why. More ways to get feedback from your first users, such as a one-question survey on the page where people leave, sit in their own guide.

If you run a D2C store: In Shopify, go to Orders › Abandoned checkouts and open each one to see the shopper's email. Send them the second message, not a discount.

• Tool: email, or the app where you met each person.
• Time: 1 hour today, then about 15 minutes a day.
• You will have: 20 messages sent today, and every answer logged in the person's words.
• It worked if: at least 6 of the 20 answer, and you can name the place, the person or the moment that brought your first customers.
• Common mistake: asking "Any feedback?". It gets polite praise. Ask about one moment: where they heard, or what stopped them.

5. Send 100 personal messages a day to new people like your first customers

Restart the work at a volume that shows you something within 4 weeks: send 100 personal messages a day to new people who look like your first customers, give 20 answers a day across 5 places like the one your first customers came from, and post 2 times a day on the channel your buyers use most. A stall at this size is usually too few new people hearing about you. More people, aimed well, is the fix.

Who to message. Use step 3 and step 4. Write down what your first customers have in common: their kind of business or life, their size, and the moment the problem hurt. That is your ideal customer profile. Then find 100 new people who match it every day. Look where your first customers were: the same kind of thread, group, job title or event. A free lead list needs about 3,000 names a month at this volume, so add 100 new names every day.

How it is possible. Give ChatGPT or Claude 10 names at a time, each with one line you found about the person, and ask for one short draft each. Read each one, make the first line true and about them, then send it. That is about a minute a message. Follow up with everyone who has not replied, twice more, 3 days apart. Answers in communities go out in your own words: AI can find the questions and list the points, and you write the reply.

Stay inside each platform's limits by spreading the volume, not cutting it. If you email more than about 50 strangers a day, set up and warm up a second domain first, so your main email stays safe. LinkedIn does not publish its limit on connection requests, and going over it usually blocks new requests for about a week, so send part of the 100 as LinkedIn messages to people you are already connected to, and the rest by email. In communities, read each place's rules on self-promotion first and answer the question before anything else, because accounts that only promote get banned. Twenty answers spread over 5 places stays well inside each one's rules.

The full daily routine for your kind of business has its own guide:

If you run a SaaS: follow your first 100 SaaS users: offer a 15-minute set-up call in every message, and set up every signup within 24 hours.

If you run a D2C store: follow your first sales on Shopify: message 100 small creators a day as well as shoppers, and post 2 short videos a day.

If you run a B2B company: follow your first B2B clients: ask each happy client for one introduction, then 100 named buyers a day, and quote the same day.

If you run an agency: follow your first agency client: message owners of the one kind of business you chose to serve, and ask each current client for one introduction.

For your 2 posts a day, pick the channel where your buyers already spend time, such as LinkedIn for business buyers. Tag the link in every post and message with a UTM tag, so step 6 can see which one worked. Leave a second launch and paid ads until this is running: one day of attention will not replace a daily source.

• Tool: your email, LinkedIn or the app your buyers use, the search box in each community, Google Sheets to log every message, and ChatGPT or Claude to draft.
• Time: about 3 to 4 hours a day, every day: 2 hours for 100 messages and follow-ups, 1 hour for 20 answers, 30 minutes for 2 posts, and 30 minutes to find 100 new names.
• You will have: 700 messages sent and logged by the end of week 1, and 2,800 by the end of week 4, plus 560 answers.
• It worked if: you get about 3 replies or first steps for every 100 messages, and by 2,800 messages your weekly count is back to at least the weekly average of your best 4 weeks. Under 2 replies per 100, rewrite the first line so it is more about the person, then send the next 100.
• Common mistake: sending one copied message to hundreds of people, or moving to a new channel where nobody like your first customers spends time. The other mistake is stopping at 200 messages because the first days were quiet.

6. Every Sunday, compare your week with your best 4 weeks and move your hours

Every Sunday, add one row to your sheet: what you did this week, what came in, and from which source. Put it next to the weekly average of your best 4 weeks. After 2,800 messages, keep what brought customers back and move the hours from anything that brought nothing.

Week | Messages sent | Answers given | Posts | Replies | First steps (signups, carts, calls) | New customers | Source of each new customer
Best 4 weeks, weekly average: [new customers per week]

Judge by how much you did, not by how many days passed. After 2,800 messages, about 4 weeks:

• Your weekly count is back to the average of your best 4 weeks: keep the same volume, and move hours from the source that brought the least to the source that brought the most.
• You get fewer than 2 replies per 100 messages: the people or the first line are wrong. Go back to step 4's answers, change who you message, and test a new first line on the next 700.
• Replies are fine, but few people take a first step: the problem is now on your site or in your offer. Fix the page for a SaaS with traffic but no signups, a store with traffic but no sales or a B2B site with no enquiries.

Once one source works again, add a second one, so the next time a source runs dry you do not start from zero. Ask every happy customer for one introduction, or start a small referral program once you have 10 or more. When you know what each source costs, work out your cost per customer by channel. Once a month, step back for a marketing review. And keep the hours in your calendar: block the outreach time in your week before you block time to build.

• Tool: Google Sheets.
• Time: 20 minutes each Sunday.
• You will have: one row a week, and a keep or move note for each source.
• It worked if: you can name the source of every new customer, and the count is at or above your best 4 weeks for 2 Sundays in a row.
• Common mistake: stopping the daily messages as soon as the count comes back. That is how most early stalls start.

Your checklist to get growth moving again

0 of 13 done

Work through this list with an AI assistant

It drafts the work and keeps track of what is done.

Do this with ChatGPT(opens in a new tab)Do this with Claude(opens in a new tab)

Step 1 · Count the last 8 weeks

• Count your signups, orders or enquiries for each of the last 8 weeks from your own records.
• Compare the first 4 weeks with the last 4 weeks and write down whether growth stalled.

Step 2 · Test it as a stranger

• Sign up, buy or enquire as a stranger on your phone and a laptop, and fix anything that fails.
• List every change to your site, product, price and email in the last 8 weeks, with dates.

Step 3 · Find the cause

• Write down where each customer came from and how much marketing work you did each week.
• Find your cause in the step 3 table.

Step 4 · Ask customers and people who stopped

• Ask 10 recent customers what brought them and 10 people who stopped what stopped them.
• Every dayAsk each new person who stops what stopped them, within 24 hours.

Step 5 · Send 100 messages a day

• Every daySend 100 personal messages to new people like your first customers.
• Every dayGive 20 answers across 5 places like the one your first customers came from.
• Every dayPost 2 times on the channel your buyers use most.

Step 6 · Compare with your best 4 weeks

• Every SundayFill in this week's row and compare it with your best 4 weeks.
• After 2,800 messagesKeep what brought customers back and move the hours from what brought nothing.

Copy listDownloadReset