Restaurant Marketing Playbook: Local SEO and Social for NZ Hospitality.

Read this first: what this page is

This is our playbook for marketing a restaurant or cafe in New Zealand. It is not a client case study. Every figure here is a sector benchmark or a planning range, not a result we are claiming for any venue.

It is deliberately cuisine-neutral, because the mechanics are the same whether you are running a bistro in Ponsonby, a brunch spot in Takapuna or a family restaurant in Manukau. Where cuisine matters, it is in the content, and we will get to that.

The real problem: you are not being found at the deciding moment

Think about how someone in Newmarket decides where to eat on a Thursday night. They open Maps or they type something like dinner near me. They see a handful of venues with a star rating, a price band, a photo and a distance. They compare those tiles for maybe fifteen seconds and pick one.

Your website plays almost no part in that. Your Google Business Profile plays nearly all of it. Which is awkward, because for most venues the profile is the least-maintained asset they own: photos from the fit-out three years ago, hours that go stale every public holiday, no menu link, and a stack of reviews nobody has replied to.

The second problem is that the audience discovering venues on social has moved almost entirely to short video. A grid of nicely lit plated dishes does not travel the way a fifteen-second clip of a room at full noise does. Plenty of NZ venues are still producing the former and wondering why reach has flattened.

The third problem is economic. Delivery apps like Uber Eats bring volume, but the commission is heavy enough that a venue can grow its order count and shrink its profit at the same time. They belong in the mix as a discovery channel, not as the strategy.

Step one: fix the Google Business Profile properly

This is unglamorous and it is the highest-return work available to a venue. Do it before anything else.

• Category. Get the primary category exactly right, then add secondary ones. A cafe that also does dinner service should say so, because the categories decide which searches you are eligible for at all.
• Photos. Upload fresh ones monthly. Interior, exterior at night so people can recognise the frontage, the food, and the room with people in it. Venues with more recent photos consistently get more profile interactions than venues with better food and older photos.
• Hours, including holidays. New Zealand has enough public holidays and regional anniversary days that special hours need a diary reminder. Nothing kills goodwill like a customer driving to a closed venue.
• Attributes. Fill in every one that applies: outdoor seating, wheelchair accessible, dog friendly, vegetarian and vegan options, gluten free, licensed, BYO, kid friendly, accepts bookings. These feed filtered searches and are pure free qualification.
• Menu. Add it as a link and as items where the format allows. Menu item text is searchable and it is how you get found for a specific dish.
• Products and services. Underused. Set piece offerings like set menus, function packages or catering deserve their own entries.

Then post to the profile roughly weekly. Specials, a new menu, an event. Posts are a modest ranking and engagement signal and they cost you five minutes.

Step two: build a review habit, not a review campaign

The realistic competitive band in New Zealand is a 4.3 to 4.6 rating. Below about 4.0 you are visibly the worse option in the map pack. A suspiciously perfect 5.0 with forty reviews reads as fake to a lot of people, so chasing that is not the goal either.

What matters more than the number is recency and volume. Twenty reviews from the last three months does more for you than two hundred from four years ago.

Getting reviews without being annoying

Ask once, at the right moment, in one channel. A short link or QR on the receipt or the table card, framed as a request from the person who served them. Train the floor team to ask verbally when a table has obviously had a good time, because a verbal ask followed by a QR converts far better than a QR alone.

Our planning benchmark is 3-8% of prompted diners leaving a review. If you serve 800 covers a week and prompt properly, that is a meaningful and steady stream. Do not buy reviews, and do not offer a discount in exchange for one. It breaches Google's policies, it is a Fair Trading Act problem in New Zealand, and diners are good at spotting it.

Replying

Reply to everything within a couple of days. Positive reviews get something short and specific. Negative reviews get an acknowledgement, an apology where one is owed, and an offer to sort it out offline. Never argue and never blame the customer in public.

The audience for a reply to a bad review is not the reviewer. It is the next hundred people reading it while deciding whether to book. A calm, human reply to a harsh review often does more good than the review did harm.

Step three: a content cadence you can actually sustain

Every venue we talk to has tried posting daily for two weeks and then stopped. Three or four posts a week, forever, beats daily for a fortnight.

Weight it towards short video. What performs for NZ hospitality, roughly in order:

• Food being made. Close, fast, sound on. The pour, the sear, the cut. It is the most reliable format there is.
• The room in service. A busy venue is social proof. Ten seconds of a full room on a Friday sells better than any still.
• People. Your chef explaining one dish, a barista, a supplier. Faces build the parasocial pull that makes someone choose you over an identical venue down the road.
• Reactive local content. Weather, a long weekend, a big match, a nearby event. Timely beats polished.
• Plated hero shots. Still worth doing, but they anchor the grid rather than drive reach.

You do not need a production company for this. A current phone, a cheap clip-on light and someone on the team who enjoys it will outperform a quarterly professional shoot, because it can be frequent. Book a photographer once or twice a year for menu and profile stills, and shoot the rest in-house.

Step four: paid social with a job to do

Do not buy followers. Buy occupancy in your quiet sessions.

Pick the specific gaps you want to fill, say Tuesday to Thursday dinner or the mid-afternoon lull, and run a tightly geo-fenced campaign inside about three to five kilometres of the venue with an offer that is genuinely worth crossing town for and has an expiry. Then make it redeemable in a way you can count, which usually means a code at the till or a specific booking type.

Our planning range for cost per tracked visit is NZD $4-$12. Work out your contribution margin per cover before you start. If the average table is worth $70 and your margin on that is $25, paying $8 to fill a seat that would otherwise be empty is obviously good, and paying $20 is not.

What this costs in NZD

Planning ranges from the New Zealand market, not quotes.

• Paid social: NZD $300-$1,500 a month per venue. Below about $200 there is not enough delivery to learn anything.
• Content and social management: NZD $800-$2,500 a month depending on how much you shoot in-house.
• Photography: NZD $600-$1,800 per session, twice a year is usually plenty.
• Local SEO and profile management: NZD $400-$1,200 a month, less for a single site.
• Website: a hospitality site that handles menus, bookings and mobile properly generally runs NZD $4,000-$12,000.

What to actually measure

• Google Business Profile interactions. Direction requests, calls and website clicks from the profile. The closest free proxy for real local demand.
• Review volume and recency. Reviews per month, not lifetime total.
• Direct bookings versus platform bookings. If direct is growing as a share, you are winning the margin battle.
• Cost per tracked visit on paid campaigns, against your contribution margin per cover.
• Covers in the sessions you targeted. Your POS already knows this. Compare like-for-like weeknights before and after.

Ignore follower count as a headline metric. It is the easiest number to move and the least connected to your till.

Common mistakes

• Neglecting the profile while polishing the website. Backwards for hospitality. The profile is the shopfront.
• Ignoring reviews, then panicking. A quiet steady habit beats a frantic push after a bad month.
• Discounting into the ground. Permanent offers train customers to never pay full price. Keep them time-boxed and tied to specific quiet sessions.
• Leaning entirely on delivery apps. Fine for discovery, dangerous as the whole strategy once you count the commission.
• Posting only finished plates. Process, people and full rooms travel much further.
• Stopping in the quiet season. Consistency across a full year is what compounds.

How long it actually takes

Month one is foundations: profile rebuilt, review system installed, content plan and shooting rhythm agreed. Expect very little visible movement.

Months two and three are usually when review volume and rating start to shift and profile interactions climb, because Maps rewards recency fairly quickly.

Months four to six are when social reach compounds and you should be able to see a difference in the specific sessions you targeted, in your own POS data rather than in a dashboard.

Months six to nine are when it becomes a habit that runs with less input, and when you can honestly compare a full quarter against the same quarter last year. Hospitality is seasonal enough in New Zealand that anything shorter than a year-on-year comparison is guesswork.